From 07eeda8ae3e1c7bdfe3fcc51a8786f35682f4c3f Mon Sep 17 00:00:00 2001 From: "stainless-app[bot]" <142633134+stainless-app[bot]@users.noreply.github.com> Date: Tue, 10 Feb 2026 22:15:42 +0000 Subject: [PATCH 1/8] chore(internal): codegen related update From bf5fa9f93b6443ac0b41d5fa134aab8ee12cfbc3 Mon Sep 17 00:00:00 2001 From: "stainless-app[bot]" <142633134+stainless-app[bot]@users.noreply.github.com> Date: Wed, 11 Feb 2026 18:33:37 +0000 Subject: [PATCH 2/8] fix(client): loosen auth header validation --- src/gradient/_client.py | 52 +-- .../knowledge_bases/knowledge_bases.py | 48 +-- .../types/retrieve_documents_params.py | 1 - tests/test_client.py | 322 +++++------------- 4 files changed, 105 insertions(+), 318 deletions(-) diff --git a/src/gradient/_client.py b/src/gradient/_client.py index 847121b1..fc58b291 100644 --- a/src/gradient/_client.py +++ b/src/gradient/_client.py @@ -353,14 +353,10 @@ def copy( Create a new client instance re-using the same options given to the current client with optional overriding. """ if default_headers is not None and set_default_headers is not None: - raise ValueError( - "The `default_headers` and `set_default_headers` arguments are mutually exclusive" - ) + raise ValueError("The `default_headers` and `set_default_headers` arguments are mutually exclusive") if default_query is not None and set_default_query is not None: - raise ValueError( - "The `default_query` and `set_default_query` arguments are mutually exclusive" - ) + raise ValueError("The `default_query` and `set_default_query` arguments are mutually exclusive") headers = self._custom_headers if default_headers is not None: @@ -411,14 +407,10 @@ def _make_status_error( return _exceptions.BadRequestError(err_msg, response=response, body=body) if response.status_code == 401: - return _exceptions.AuthenticationError( - err_msg, response=response, body=body - ) + return _exceptions.AuthenticationError(err_msg, response=response, body=body) if response.status_code == 403: - return _exceptions.PermissionDeniedError( - err_msg, response=response, body=body - ) + return _exceptions.PermissionDeniedError(err_msg, response=response, body=body) if response.status_code == 404: return _exceptions.NotFoundError(err_msg, response=response, body=body) @@ -427,17 +419,13 @@ def _make_status_error( return _exceptions.ConflictError(err_msg, response=response, body=body) if response.status_code == 422: - return _exceptions.UnprocessableEntityError( - err_msg, response=response, body=body - ) + return _exceptions.UnprocessableEntityError(err_msg, response=response, body=body) if response.status_code == 429: return _exceptions.RateLimitError(err_msg, response=response, body=body) if response.status_code >= 500: - return _exceptions.InternalServerError( - err_msg, response=response, body=body - ) + return _exceptions.InternalServerError(err_msg, response=response, body=body) return APIStatusError(err_msg, response=response, body=body) @@ -719,14 +707,10 @@ def copy( Create a new client instance re-using the same options given to the current client with optional overriding. """ if default_headers is not None and set_default_headers is not None: - raise ValueError( - "The `default_headers` and `set_default_headers` arguments are mutually exclusive" - ) + raise ValueError("The `default_headers` and `set_default_headers` arguments are mutually exclusive") if default_query is not None and set_default_query is not None: - raise ValueError( - "The `default_query` and `set_default_query` arguments are mutually exclusive" - ) + raise ValueError("The `default_query` and `set_default_query` arguments are mutually exclusive") headers = self._custom_headers if default_headers is not None: @@ -777,14 +761,10 @@ def _make_status_error( return _exceptions.BadRequestError(err_msg, response=response, body=body) if response.status_code == 401: - return _exceptions.AuthenticationError( - err_msg, response=response, body=body - ) + return _exceptions.AuthenticationError(err_msg, response=response, body=body) if response.status_code == 403: - return _exceptions.PermissionDeniedError( - err_msg, response=response, body=body - ) + return _exceptions.PermissionDeniedError(err_msg, response=response, body=body) if response.status_code == 404: return _exceptions.NotFoundError(err_msg, response=response, body=body) @@ -793,17 +773,13 @@ def _make_status_error( return _exceptions.ConflictError(err_msg, response=response, body=body) if response.status_code == 422: - return _exceptions.UnprocessableEntityError( - err_msg, response=response, body=body - ) + return _exceptions.UnprocessableEntityError(err_msg, response=response, body=body) if response.status_code == 429: return _exceptions.RateLimitError(err_msg, response=response, body=body) if response.status_code >= 500: - return _exceptions.InternalServerError( - err_msg, response=response, body=body - ) + return _exceptions.InternalServerError(err_msg, response=response, body=body) return APIStatusError(err_msg, response=response, body=body) @@ -1082,9 +1058,7 @@ def knowledge_bases( AsyncKnowledgeBasesResourceWithStreamingResponse, ) - return AsyncKnowledgeBasesResourceWithStreamingResponse( - self._client.knowledge_bases - ) + return AsyncKnowledgeBasesResourceWithStreamingResponse(self._client.knowledge_bases) @cached_property def models(self) -> models.AsyncModelsResourceWithStreamingResponse: diff --git a/src/gradient/resources/knowledge_bases/knowledge_bases.py b/src/gradient/resources/knowledge_bases/knowledge_bases.py index 04c19b32..9c22a60c 100644 --- a/src/gradient/resources/knowledge_bases/knowledge_bases.py +++ b/src/gradient/resources/knowledge_bases/knowledge_bases.py @@ -210,9 +210,7 @@ def retrieve( timeout: Override the client-level default timeout for this request, in seconds """ if not uuid: - raise ValueError( - f"Expected a non-empty value for `uuid` but received {uuid!r}" - ) + raise ValueError(f"Expected a non-empty value for `uuid` but received {uuid!r}") return self._get( ( f"/v2/gen-ai/knowledge_bases/{uuid}" @@ -271,9 +269,7 @@ def update( timeout: Override the client-level default timeout for this request, in seconds """ if not path_uuid: - raise ValueError( - f"Expected a non-empty value for `path_uuid` but received {path_uuid!r}" - ) + raise ValueError(f"Expected a non-empty value for `path_uuid` but received {path_uuid!r}") return self._put( ( f"/v2/gen-ai/knowledge_bases/{path_uuid}" @@ -375,9 +371,7 @@ def delete( timeout: Override the client-level default timeout for this request, in seconds """ if not uuid: - raise ValueError( - f"Expected a non-empty value for `uuid` but received {uuid!r}" - ) + raise ValueError(f"Expected a non-empty value for `uuid` but received {uuid!r}") return self._delete( ( f"/v2/gen-ai/knowledge_bases/{uuid}" @@ -646,9 +640,7 @@ async def retrieve( timeout: Override the client-level default timeout for this request, in seconds """ if not uuid: - raise ValueError( - f"Expected a non-empty value for `uuid` but received {uuid!r}" - ) + raise ValueError(f"Expected a non-empty value for `uuid` but received {uuid!r}") return await self._get( ( f"/v2/gen-ai/knowledge_bases/{uuid}" @@ -707,9 +699,7 @@ async def update( timeout: Override the client-level default timeout for this request, in seconds """ if not path_uuid: - raise ValueError( - f"Expected a non-empty value for `path_uuid` but received {path_uuid!r}" - ) + raise ValueError(f"Expected a non-empty value for `path_uuid` but received {path_uuid!r}") return await self._put( ( f"/v2/gen-ai/knowledge_bases/{path_uuid}" @@ -811,9 +801,7 @@ async def delete( timeout: Override the client-level default timeout for this request, in seconds """ if not uuid: - raise ValueError( - f"Expected a non-empty value for `uuid` but received {uuid!r}" - ) + raise ValueError(f"Expected a non-empty value for `uuid` but received {uuid!r}") return await self._delete( ( f"/v2/gen-ai/knowledge_bases/{uuid}" @@ -1011,15 +999,11 @@ def __init__(self, knowledge_bases: AsyncKnowledgeBasesResource) -> None: @cached_property def data_sources(self) -> AsyncDataSourcesResourceWithRawResponse: - return AsyncDataSourcesResourceWithRawResponse( - self._knowledge_bases.data_sources - ) + return AsyncDataSourcesResourceWithRawResponse(self._knowledge_bases.data_sources) @cached_property def indexing_jobs(self) -> AsyncIndexingJobsResourceWithRawResponse: - return AsyncIndexingJobsResourceWithRawResponse( - self._knowledge_bases.indexing_jobs - ) + return AsyncIndexingJobsResourceWithRawResponse(self._knowledge_bases.indexing_jobs) class KnowledgeBasesResourceWithStreamingResponse: @@ -1050,15 +1034,11 @@ def __init__(self, knowledge_bases: KnowledgeBasesResource) -> None: @cached_property def data_sources(self) -> DataSourcesResourceWithStreamingResponse: - return DataSourcesResourceWithStreamingResponse( - self._knowledge_bases.data_sources - ) + return DataSourcesResourceWithStreamingResponse(self._knowledge_bases.data_sources) @cached_property def indexing_jobs(self) -> IndexingJobsResourceWithStreamingResponse: - return IndexingJobsResourceWithStreamingResponse( - self._knowledge_bases.indexing_jobs - ) + return IndexingJobsResourceWithStreamingResponse(self._knowledge_bases.indexing_jobs) class AsyncKnowledgeBasesResourceWithStreamingResponse: @@ -1089,12 +1069,8 @@ def __init__(self, knowledge_bases: AsyncKnowledgeBasesResource) -> None: @cached_property def data_sources(self) -> AsyncDataSourcesResourceWithStreamingResponse: - return AsyncDataSourcesResourceWithStreamingResponse( - self._knowledge_bases.data_sources - ) + return AsyncDataSourcesResourceWithStreamingResponse(self._knowledge_bases.data_sources) @cached_property def indexing_jobs(self) -> AsyncIndexingJobsResourceWithStreamingResponse: - return AsyncIndexingJobsResourceWithStreamingResponse( - self._knowledge_bases.indexing_jobs - ) + return AsyncIndexingJobsResourceWithStreamingResponse(self._knowledge_bases.indexing_jobs) diff --git a/src/gradient/types/retrieve_documents_params.py b/src/gradient/types/retrieve_documents_params.py index bad99ad3..359d8a07 100644 --- a/src/gradient/types/retrieve_documents_params.py +++ b/src/gradient/types/retrieve_documents_params.py @@ -73,4 +73,3 @@ class Filters(TypedDict, total=False): should: Iterable[FiltersShould] """At least one condition must match (OR)""" - diff --git a/tests/test_client.py b/tests/test_client.py index 4b645c08..c78fa9f8 100644 --- a/tests/test_client.py +++ b/tests/test_client.py @@ -59,9 +59,7 @@ def _low_retry_timeout(*_args: Any, **_kwargs: Any) -> float: def _get_open_connections(client: Gradient | AsyncGradient) -> int: transport = client._client._transport - assert isinstance(transport, httpx.HTTPTransport) or isinstance( - transport, httpx.AsyncHTTPTransport - ) + assert isinstance(transport, httpx.HTTPTransport) or isinstance(transport, httpx.AsyncHTTPTransport) pool = transport._pool return len(pool._requests) @@ -70,9 +68,7 @@ def _get_open_connections(client: Gradient | AsyncGradient) -> int: class TestGradient: @pytest.mark.respx(base_url=base_url) def test_raw_response(self, respx_mock: MockRouter, client: Gradient) -> None: - respx_mock.post("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.post("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) response = client.post("/foo", cast_to=httpx.Response) assert response.status_code == 200 @@ -80,9 +76,7 @@ def test_raw_response(self, respx_mock: MockRouter, client: Gradient) -> None: assert response.json() == {"foo": "bar"} @pytest.mark.respx(base_url=base_url) - def test_raw_response_for_binary( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_raw_response_for_binary(self, respx_mock: MockRouter, client: Gradient) -> None: respx_mock.post("/foo").mock( return_value=httpx.Response( 200, @@ -231,9 +225,7 @@ def test_copy_signature(self, client: Gradient) -> None: continue copy_param = copy_signature.parameters.get(name) - assert ( - copy_param is not None - ), f"copy() signature is missing the {name} param" + assert copy_param is not None, f"copy() signature is missing the {name} param" @pytest.mark.skipif( sys.version_info >= (3, 10), @@ -269,9 +261,7 @@ def build_request(options: FinalRequestOptions) -> None: tracemalloc.stop() - def add_leak( - leaks: list[tracemalloc.StatisticDiff], diff: tracemalloc.StatisticDiff - ) -> None: + def add_leak(leaks: list[tracemalloc.StatisticDiff], diff: tracemalloc.StatisticDiff) -> None: if diff.count == 0: # Avoid false positives by considering only leaks (i.e. allocations that persist). return @@ -314,9 +304,7 @@ def test_request_timeout(self, client: Gradient) -> None: timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == DEFAULT_TIMEOUT - request = client._build_request( - FinalRequestOptions(method="get", url="/foo", timeout=httpx.Timeout(100.0)) - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo", timeout=httpx.Timeout(100.0))) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == httpx.Timeout(100.0) @@ -348,9 +336,7 @@ def test_http_client_timeout_option(self) -> None: http_client=http_client, ) - request = client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == httpx.Timeout(None) @@ -367,9 +353,7 @@ def test_http_client_timeout_option(self) -> None: http_client=http_client, ) - request = client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == DEFAULT_TIMEOUT @@ -386,9 +370,7 @@ def test_http_client_timeout_option(self) -> None: http_client=http_client, ) - request = client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == DEFAULT_TIMEOUT # our default @@ -415,9 +397,7 @@ def test_default_headers_option(self) -> None: _strict_response_validation=True, default_headers={"X-Foo": "bar"}, ) - request = test_client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = test_client._build_request(FinalRequestOptions(method="get", url="/foo")) assert request.headers.get("x-foo") == "bar" assert request.headers.get("x-stainless-lang") == "python" @@ -432,9 +412,7 @@ def test_default_headers_option(self) -> None: "X-Stainless-Lang": "my-overriding-header", }, ) - request = test_client2._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = test_client2._build_request(FinalRequestOptions(method="get", url="/foo")) assert request.headers.get("x-foo") == "stainless" assert request.headers.get("x-stainless-lang") == "my-overriding-header" @@ -474,9 +452,7 @@ def test_validate_headers(self) -> None: client2._build_request(FinalRequestOptions(method="get", url="/foo")) request2 = client2._build_request( - FinalRequestOptions( - method="get", url="/foo", headers={"Authorization": Omit()} - ) + FinalRequestOptions(method="get", url="/foo", headers={"Authorization": Omit()}) ) assert request2.headers.get("Authorization") is None @@ -607,9 +583,7 @@ def test_multipart_repeating_array(self, client: Gradient) -> None: FinalRequestOptions.construct( method="post", url="/foo", - headers={ - "Content-Type": "multipart/form-data; boundary=6b7ba517decee4a450543ea6ae821c82" - }, + headers={"Content-Type": "multipart/form-data; boundary=6b7ba517decee4a450543ea6ae821c82"}, json_data={"array": ["foo", "bar"]}, files=[("foo.txt", b"hello world")], ) @@ -634,27 +608,21 @@ def test_multipart_repeating_array(self, client: Gradient) -> None: ] @pytest.mark.respx(base_url=base_url) - def test_basic_union_response( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_basic_union_response(self, respx_mock: MockRouter, client: Gradient) -> None: class Model1(BaseModel): name: str class Model2(BaseModel): foo: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) response = client.get("/foo", cast_to=cast(Any, Union[Model1, Model2])) assert isinstance(response, Model2) assert response.foo == "bar" @pytest.mark.respx(base_url=base_url) - def test_union_response_different_types( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_union_response_different_types(self, respx_mock: MockRouter, client: Gradient) -> None: """Union of objects with the same field name using a different type""" class Model1(BaseModel): @@ -663,9 +631,7 @@ class Model1(BaseModel): class Model2(BaseModel): foo: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) response = client.get("/foo", cast_to=cast(Any, Union[Model1, Model2])) assert isinstance(response, Model2) @@ -678,9 +644,7 @@ class Model2(BaseModel): assert response.foo == 1 @pytest.mark.respx(base_url=base_url) - def test_non_application_json_content_type_for_json_data( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_non_application_json_content_type_for_json_data(self, respx_mock: MockRouter, client: Gradient) -> None: """ Response that sets Content-Type to something other than application/json but returns json data """ @@ -850,15 +814,11 @@ def test_client_context_manager(self) -> None: assert test_client.is_closed() @pytest.mark.respx(base_url=base_url) - def test_client_response_validation_error( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_client_response_validation_error(self, respx_mock: MockRouter, client: Gradient) -> None: class Model(BaseModel): foo: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, json={"foo": {"invalid": True}}) - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": {"invalid": True}})) with pytest.raises(APIResponseValidationError) as exc: client.get("/foo", cast_to=Model) @@ -881,13 +841,9 @@ def test_default_stream_cls(self, respx_mock: MockRouter, client: Gradient) -> N class Model(BaseModel): name: str - respx_mock.post("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.post("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) - stream = client.post( - "/foo", cast_to=Model, stream=True, stream_cls=Stream[Model] - ) + stream = client.post("/foo", cast_to=Model, stream=True, stream_cls=Stream[Model]) assert isinstance(stream, Stream) stream.response.close() @@ -896,9 +852,7 @@ def test_received_text_for_expected_json(self, respx_mock: MockRouter) -> None: class Model(BaseModel): name: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, text="my-custom-format") - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, text="my-custom-format")) strict_client = Gradient( base_url=base_url, @@ -959,16 +913,10 @@ def test_parse_retry_after_header(self, remaining_retries: int, retry_after: str calculated = client._calculate_retry_timeout(remaining_retries, options, headers) assert calculated == pytest.approx(timeout, rel=0.5 * 0.875) # type: ignore[misc] - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) - def test_retrying_timeout_errors_doesnt_leak( - self, respx_mock: MockRouter, client: Gradient - ) -> None: - respx_mock.post("/chat/completions").mock( - side_effect=httpx.TimeoutException("Test timeout error") - ) + def test_retrying_timeout_errors_doesnt_leak(self, respx_mock: MockRouter, client: Gradient) -> None: + respx_mock.post("/chat/completions").mock(side_effect=httpx.TimeoutException("Test timeout error")) with pytest.raises(APITimeoutError): client.chat.completions.with_streaming_response.create( @@ -983,13 +931,9 @@ def test_retrying_timeout_errors_doesnt_leak( assert _get_open_connections(client) == 0 - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) - def test_retrying_status_errors_doesnt_leak( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_retrying_status_errors_doesnt_leak(self, respx_mock: MockRouter, client: Gradient) -> None: respx_mock.post("/chat/completions").mock(return_value=httpx.Response(500)) with pytest.raises(APIStatusError): @@ -1005,9 +949,7 @@ def test_retrying_status_errors_doesnt_leak( assert _get_open_connections(client) == 0 @pytest.mark.parametrize("failures_before_success", [0, 2, 4]) - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) @pytest.mark.parametrize("failure_mode", ["status", "exception"]) def test_retries_taken( @@ -1043,15 +985,10 @@ def retry_handler(_request: httpx.Request) -> httpx.Response: ) assert response.retries_taken == failures_before_success - assert ( - int(response.http_request.headers.get("x-stainless-retry-count")) - == failures_before_success - ) + assert int(response.http_request.headers.get("x-stainless-retry-count")) == failures_before_success @pytest.mark.parametrize("failures_before_success", [0, 2, 4]) - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) def test_omit_retry_count_header( self, client: Gradient, failures_before_success: int, respx_mock: MockRouter @@ -1080,14 +1017,10 @@ def retry_handler(_request: httpx.Request) -> httpx.Response: extra_headers={"x-stainless-retry-count": Omit()}, ) - assert ( - len(response.http_request.headers.get_list("x-stainless-retry-count")) == 0 - ) + assert len(response.http_request.headers.get_list("x-stainless-retry-count")) == 0 @pytest.mark.parametrize("failures_before_success", [0, 2, 4]) - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) def test_overwrite_retry_count_header( self, client: Gradient, failures_before_success: int, respx_mock: MockRouter @@ -1144,29 +1077,19 @@ def test_default_client_creation(self) -> None: def test_follow_redirects(self, respx_mock: MockRouter, client: Gradient) -> None: # Test that the default follow_redirects=True allows following redirects respx_mock.post("/redirect").mock( - return_value=httpx.Response( - 302, headers={"Location": f"{base_url}/redirected"} - ) - ) - respx_mock.get("/redirected").mock( - return_value=httpx.Response(200, json={"status": "ok"}) + return_value=httpx.Response(302, headers={"Location": f"{base_url}/redirected"}) ) + respx_mock.get("/redirected").mock(return_value=httpx.Response(200, json={"status": "ok"})) - response = client.post( - "/redirect", body={"key": "value"}, cast_to=httpx.Response - ) + response = client.post("/redirect", body={"key": "value"}, cast_to=httpx.Response) assert response.status_code == 200 assert response.json() == {"status": "ok"} @pytest.mark.respx(base_url=base_url) - def test_follow_redirects_disabled( - self, respx_mock: MockRouter, client: Gradient - ) -> None: + def test_follow_redirects_disabled(self, respx_mock: MockRouter, client: Gradient) -> None: # Test that follow_redirects=False prevents following redirects respx_mock.post("/redirect").mock( - return_value=httpx.Response( - 302, headers={"Location": f"{base_url}/redirected"} - ) + return_value=httpx.Response(302, headers={"Location": f"{base_url}/redirected"}) ) with pytest.raises(APIStatusError) as exc_info: @@ -1183,12 +1106,8 @@ def test_follow_redirects_disabled( class TestAsyncGradient: @pytest.mark.respx(base_url=base_url) - async def test_raw_response( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: - respx_mock.post("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + async def test_raw_response(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: + respx_mock.post("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) response = await async_client.post("/foo", cast_to=httpx.Response) assert response.status_code == 200 @@ -1196,9 +1115,7 @@ async def test_raw_response( assert response.json() == {"foo": "bar"} @pytest.mark.respx(base_url=base_url) - async def test_raw_response_for_binary( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_raw_response_for_binary(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: respx_mock.post("/foo").mock( return_value=httpx.Response( 200, @@ -1347,9 +1264,7 @@ def test_copy_signature(self, async_client: AsyncGradient) -> None: continue copy_param = copy_signature.parameters.get(name) - assert ( - copy_param is not None - ), f"copy() signature is missing the {name} param" + assert copy_param is not None, f"copy() signature is missing the {name} param" @pytest.mark.skipif( sys.version_info >= (3, 10), @@ -1385,9 +1300,7 @@ def build_request(options: FinalRequestOptions) -> None: tracemalloc.stop() - def add_leak( - leaks: list[tracemalloc.StatisticDiff], diff: tracemalloc.StatisticDiff - ) -> None: + def add_leak(leaks: list[tracemalloc.StatisticDiff], diff: tracemalloc.StatisticDiff) -> None: if diff.count == 0: # Avoid false positives by considering only leaks (i.e. allocations that persist). return @@ -1426,9 +1339,7 @@ def add_leak( raise AssertionError() async def test_request_timeout(self, async_client: AsyncGradient) -> None: - request = async_client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = async_client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == DEFAULT_TIMEOUT @@ -1466,9 +1377,7 @@ async def test_http_client_timeout_option(self) -> None: http_client=http_client, ) - request = client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == httpx.Timeout(None) @@ -1485,9 +1394,7 @@ async def test_http_client_timeout_option(self) -> None: http_client=http_client, ) - request = client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == DEFAULT_TIMEOUT @@ -1504,9 +1411,7 @@ async def test_http_client_timeout_option(self) -> None: http_client=http_client, ) - request = client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = client._build_request(FinalRequestOptions(method="get", url="/foo")) timeout = httpx.Timeout(**request.extensions["timeout"]) # type: ignore assert timeout == DEFAULT_TIMEOUT # our default @@ -1533,9 +1438,7 @@ async def test_default_headers_option(self) -> None: _strict_response_validation=True, default_headers={"X-Foo": "bar"}, ) - request = test_client._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = test_client._build_request(FinalRequestOptions(method="get", url="/foo")) assert request.headers.get("x-foo") == "bar" assert request.headers.get("x-stainless-lang") == "python" @@ -1550,9 +1453,7 @@ async def test_default_headers_option(self) -> None: "X-Stainless-Lang": "my-overriding-header", }, ) - request = test_client2._build_request( - FinalRequestOptions(method="get", url="/foo") - ) + request = test_client2._build_request(FinalRequestOptions(method="get", url="/foo")) assert request.headers.get("x-foo") == "stainless" assert request.headers.get("x-stainless-lang") == "my-overriding-header" @@ -1592,9 +1493,7 @@ def test_validate_headers(self) -> None: client2._build_request(FinalRequestOptions(method="get", url="/foo")) request2 = client2._build_request( - FinalRequestOptions( - method="get", url="/foo", headers={"Authorization": Omit()} - ) + FinalRequestOptions(method="get", url="/foo", headers={"Authorization": Omit()}) ) assert request2.headers.get("Authorization") is None @@ -1725,9 +1624,7 @@ def test_multipart_repeating_array(self, async_client: AsyncGradient) -> None: FinalRequestOptions.construct( method="post", url="/foo", - headers={ - "Content-Type": "multipart/form-data; boundary=6b7ba517decee4a450543ea6ae821c82" - }, + headers={"Content-Type": "multipart/form-data; boundary=6b7ba517decee4a450543ea6ae821c82"}, json_data={"array": ["foo", "bar"]}, files=[("foo.txt", b"hello world")], ) @@ -1752,29 +1649,21 @@ def test_multipart_repeating_array(self, async_client: AsyncGradient) -> None: ] @pytest.mark.respx(base_url=base_url) - async def test_basic_union_response( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_basic_union_response(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: class Model1(BaseModel): name: str class Model2(BaseModel): foo: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) - response = await async_client.get( - "/foo", cast_to=cast(Any, Union[Model1, Model2]) - ) + response = await async_client.get("/foo", cast_to=cast(Any, Union[Model1, Model2])) assert isinstance(response, Model2) assert response.foo == "bar" @pytest.mark.respx(base_url=base_url) - async def test_union_response_different_types( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_union_response_different_types(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: """Union of objects with the same field name using a different type""" class Model1(BaseModel): @@ -1783,21 +1672,15 @@ class Model1(BaseModel): class Model2(BaseModel): foo: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) - response = await async_client.get( - "/foo", cast_to=cast(Any, Union[Model1, Model2]) - ) + response = await async_client.get("/foo", cast_to=cast(Any, Union[Model1, Model2])) assert isinstance(response, Model2) assert response.foo == "bar" respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": 1})) - response = await async_client.get( - "/foo", cast_to=cast(Any, Union[Model1, Model2]) - ) + response = await async_client.get("/foo", cast_to=cast(Any, Union[Model1, Model2])) assert isinstance(response, Model1) assert response.foo == 1 @@ -1979,15 +1862,11 @@ async def test_client_context_manager(self) -> None: assert test_client.is_closed() @pytest.mark.respx(base_url=base_url) - async def test_client_response_validation_error( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_client_response_validation_error(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: class Model(BaseModel): foo: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, json={"foo": {"invalid": True}}) - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, json={"foo": {"invalid": True}})) with pytest.raises(APIResponseValidationError) as exc: await async_client.get("/foo", cast_to=Model) @@ -2006,32 +1885,22 @@ async def test_client_max_retries_validation(self) -> None: ) @pytest.mark.respx(base_url=base_url) - async def test_default_stream_cls( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_default_stream_cls(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: class Model(BaseModel): name: str - respx_mock.post("/foo").mock( - return_value=httpx.Response(200, json={"foo": "bar"}) - ) + respx_mock.post("/foo").mock(return_value=httpx.Response(200, json={"foo": "bar"})) - stream = await async_client.post( - "/foo", cast_to=Model, stream=True, stream_cls=AsyncStream[Model] - ) + stream = await async_client.post("/foo", cast_to=Model, stream=True, stream_cls=AsyncStream[Model]) assert isinstance(stream, AsyncStream) await stream.response.aclose() @pytest.mark.respx(base_url=base_url) - async def test_received_text_for_expected_json( - self, respx_mock: MockRouter - ) -> None: + async def test_received_text_for_expected_json(self, respx_mock: MockRouter) -> None: class Model(BaseModel): name: str - respx_mock.get("/foo").mock( - return_value=httpx.Response(200, text="my-custom-format") - ) + respx_mock.get("/foo").mock(return_value=httpx.Response(200, text="my-custom-format")) strict_client = AsyncGradient( base_url=base_url, @@ -2095,16 +1964,12 @@ async def test_parse_retry_after_header(self, remaining_retries: int, retry_afte calculated = async_client._calculate_retry_timeout(remaining_retries, options, headers) assert calculated == pytest.approx(timeout, rel=0.5 * 0.875) # type: ignore[misc] - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) async def test_retrying_timeout_errors_doesnt_leak( self, respx_mock: MockRouter, async_client: AsyncGradient ) -> None: - respx_mock.post("/chat/completions").mock( - side_effect=httpx.TimeoutException("Test timeout error") - ) + respx_mock.post("/chat/completions").mock(side_effect=httpx.TimeoutException("Test timeout error")) with pytest.raises(APITimeoutError): await async_client.chat.completions.with_streaming_response.create( @@ -2119,9 +1984,7 @@ async def test_retrying_timeout_errors_doesnt_leak( assert _get_open_connections(async_client) == 0 - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) async def test_retrying_status_errors_doesnt_leak( self, respx_mock: MockRouter, async_client: AsyncGradient @@ -2141,9 +2004,7 @@ async def test_retrying_status_errors_doesnt_leak( assert _get_open_connections(async_client) == 0 @pytest.mark.parametrize("failures_before_success", [0, 2, 4]) - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) @pytest.mark.parametrize("failure_mode", ["status", "exception"]) async def test_retries_taken( @@ -2179,15 +2040,10 @@ def retry_handler(_request: httpx.Request) -> httpx.Response: ) assert response.retries_taken == failures_before_success - assert ( - int(response.http_request.headers.get("x-stainless-retry-count")) - == failures_before_success - ) + assert int(response.http_request.headers.get("x-stainless-retry-count")) == failures_before_success @pytest.mark.parametrize("failures_before_success", [0, 2, 4]) - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) async def test_omit_retry_count_header( self, @@ -2219,14 +2075,10 @@ def retry_handler(_request: httpx.Request) -> httpx.Response: extra_headers={"x-stainless-retry-count": Omit()}, ) - assert ( - len(response.http_request.headers.get_list("x-stainless-retry-count")) == 0 - ) + assert len(response.http_request.headers.get_list("x-stainless-retry-count")) == 0 @pytest.mark.parametrize("failures_before_success", [0, 2, 4]) - @mock.patch( - "gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout - ) + @mock.patch("gradient._base_client.BaseClient._calculate_retry_timeout", _low_retry_timeout) @pytest.mark.respx(base_url=base_url) async def test_overwrite_retry_count_header( self, @@ -2264,9 +2116,7 @@ async def test_get_platform(self) -> None: platform = await asyncify(get_platform)() assert isinstance(platform, (str, OtherPlatform)) - async def test_proxy_environment_variables( - self, monkeypatch: pytest.MonkeyPatch - ) -> None: + async def test_proxy_environment_variables(self, monkeypatch: pytest.MonkeyPatch) -> None: # Test that the proxy environment variables are set correctly monkeypatch.setenv("HTTPS_PROXY", "https://example.org") @@ -2289,34 +2139,22 @@ async def test_default_client_creation(self) -> None: ) @pytest.mark.respx(base_url=base_url) - async def test_follow_redirects( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_follow_redirects(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: # Test that the default follow_redirects=True allows following redirects respx_mock.post("/redirect").mock( - return_value=httpx.Response( - 302, headers={"Location": f"{base_url}/redirected"} - ) - ) - respx_mock.get("/redirected").mock( - return_value=httpx.Response(200, json={"status": "ok"}) + return_value=httpx.Response(302, headers={"Location": f"{base_url}/redirected"}) ) + respx_mock.get("/redirected").mock(return_value=httpx.Response(200, json={"status": "ok"})) - response = await async_client.post( - "/redirect", body={"key": "value"}, cast_to=httpx.Response - ) + response = await async_client.post("/redirect", body={"key": "value"}, cast_to=httpx.Response) assert response.status_code == 200 assert response.json() == {"status": "ok"} @pytest.mark.respx(base_url=base_url) - async def test_follow_redirects_disabled( - self, respx_mock: MockRouter, async_client: AsyncGradient - ) -> None: + async def test_follow_redirects_disabled(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: # Test that follow_redirects=False prevents following redirects respx_mock.post("/redirect").mock( - return_value=httpx.Response( - 302, headers={"Location": f"{base_url}/redirected"} - ) + return_value=httpx.Response(302, headers={"Location": f"{base_url}/redirected"}) ) with pytest.raises(APIStatusError) as exc_info: From f67228f0bd8de96342d15b1b7c3096e4b03c7aaa Mon Sep 17 00:00:00 2001 From: "stainless-app[bot]" <142633134+stainless-app[bot]@users.noreply.github.com> Date: Wed, 11 Feb 2026 22:42:20 +0000 Subject: [PATCH 3/8] feat(api): api update --- .github/workflows/ci.yml | 6 +- .github/workflows/publish-pypi.yml | 2 +- .github/workflows/release-doctor.yml | 2 +- .stats.yml | 8 +- api.md | 36 + requirements-dev.lock | 20 +- requirements.lock | 8 +- src/gradient/_base_client.py | 168 +++- src/gradient/_client.py | 118 ++- src/gradient/_compat.py | 6 +- src/gradient/_models.py | 17 +- src/gradient/_types.py | 9 + src/gradient/_utils/_compat.py | 2 +- src/gradient/_utils/_json.py | 35 + src/gradient/resources/__init__.py | 42 + .../resources/agents/evaluation_datasets.py | 11 + .../resources/agents/evaluation_runs.py | 12 +- .../resources/agents/evaluation_test_cases.py | 4 + src/gradient/resources/apps/__init__.py | 33 + src/gradient/resources/apps/apps.py | 102 +++ .../resources/apps/job_invocations.py | 191 ++++ src/gradient/resources/billing.py | 226 +++++ .../resources/knowledge_bases/data_sources.py | 213 +++++ src/gradient/resources/responses.py | 860 ++++++++++++++++++ src/gradient/resources/retrieve.py | 12 + src/gradient/types/__init__.py | 7 + .../types/agents/api_evaluation_prompt.py | 57 +- .../types/agents/api_evaluation_run.py | 3 + .../agents/chat/completion_create_params.py | 315 ++++++- .../evaluation_dataset_create_params.py | 6 +- .../agents/evaluation_run_create_params.py | 5 +- .../evaluation_test_case_create_params.py | 2 + src/gradient/types/api_agent_model.py | 9 + src/gradient/types/api_model.py | 9 + src/gradient/types/apps/__init__.py | 6 + .../apps/job_invocation_cancel_params.py | 14 + .../apps/job_invocation_cancel_response.py | 77 ++ .../types/billing_list_insights_params.py | 23 + .../types/billing_list_insights_response.py | 51 ++ .../types/chat/completion_create_params.py | 315 ++++++- .../types/knowledge_base_create_params.py | 49 +- .../types/knowledge_bases/__init__.py | 2 + .../api_knowledge_base_data_source.py | 50 +- .../data_source_create_params.py | 43 +- .../data_source_update_params.py | 58 ++ .../data_source_update_response.py | 17 + src/gradient/types/response_create_params.py | 331 +++++++ src/gradient/types/shared/__init__.py | 2 + src/gradient/types/shared/completion_usage.py | 21 +- .../types/shared/create_response_response.py | 332 +++++++ .../shared/create_response_stream_response.py | 139 +++ .../agents/test_evaluation_datasets.py | 2 + .../agents/test_evaluation_runs.py | 2 + .../agents/test_evaluation_test_cases.py | 2 + tests/api_resources/apps/__init__.py | 1 + .../apps/test_job_invocations.py | 148 +++ .../api_resources/gpu_droplets/test_images.py | 8 +- .../knowledge_bases/test_data_sources.py | 159 ++++ tests/api_resources/test_billing.py | 177 ++++ tests/api_resources/test_knowledge_bases.py | 14 + tests/api_resources/test_responses.py | 296 ++++++ tests/test_client.py | 187 +++- tests/test_utils/test_json.py | 126 +++ 63 files changed, 5096 insertions(+), 112 deletions(-) create mode 100644 src/gradient/_utils/_json.py create mode 100644 src/gradient/resources/apps/__init__.py create mode 100644 src/gradient/resources/apps/apps.py create mode 100644 src/gradient/resources/apps/job_invocations.py create mode 100644 src/gradient/resources/billing.py create mode 100644 src/gradient/resources/responses.py create mode 100644 src/gradient/types/apps/__init__.py create mode 100644 src/gradient/types/apps/job_invocation_cancel_params.py create mode 100644 src/gradient/types/apps/job_invocation_cancel_response.py create mode 100644 src/gradient/types/billing_list_insights_params.py create mode 100644 src/gradient/types/billing_list_insights_response.py create mode 100644 src/gradient/types/knowledge_bases/data_source_update_params.py create mode 100644 src/gradient/types/knowledge_bases/data_source_update_response.py create mode 100644 src/gradient/types/response_create_params.py create mode 100644 src/gradient/types/shared/create_response_response.py create mode 100644 src/gradient/types/shared/create_response_stream_response.py create mode 100644 tests/api_resources/apps/__init__.py create mode 100644 tests/api_resources/apps/test_job_invocations.py create mode 100644 tests/api_resources/test_billing.py create mode 100644 tests/api_resources/test_responses.py create mode 100644 tests/test_utils/test_json.py diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml index 42a31c2a..9fb3fd00 100644 --- a/.github/workflows/ci.yml +++ b/.github/workflows/ci.yml @@ -19,7 +19,7 @@ jobs: runs-on: ${{ github.repository == 'stainless-sdks/gradient-python' && 'depot-ubuntu-24.04' || 'ubuntu-latest' }} if: github.event_name == 'push' || github.event.pull_request.head.repo.fork steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v6 - name: Install Rye run: | @@ -44,7 +44,7 @@ jobs: id-token: write runs-on: ${{ github.repository == 'stainless-sdks/gradient-python' && 'depot-ubuntu-24.04' || 'ubuntu-latest' }} steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v6 - name: Install Rye run: | @@ -81,7 +81,7 @@ jobs: runs-on: ${{ github.repository == 'stainless-sdks/gradient-python' && 'depot-ubuntu-24.04' || 'ubuntu-latest' }} if: github.event_name == 'push' || github.event.pull_request.head.repo.fork steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v6 - name: Install Rye run: | diff --git a/.github/workflows/publish-pypi.yml b/.github/workflows/publish-pypi.yml index cb2d6509..2b8edb2f 100644 --- a/.github/workflows/publish-pypi.yml +++ b/.github/workflows/publish-pypi.yml @@ -14,7 +14,7 @@ jobs: runs-on: ubuntu-latest steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v6 - name: Install Rye run: | diff --git a/.github/workflows/release-doctor.yml b/.github/workflows/release-doctor.yml index 9c8912bc..bde3b3ed 100644 --- a/.github/workflows/release-doctor.yml +++ b/.github/workflows/release-doctor.yml @@ -12,7 +12,7 @@ jobs: if: github.repository == 'digitalocean/gradient-python' && (github.event_name == 'push' || github.event_name == 'workflow_dispatch' || startsWith(github.head_ref, 'release-please') || github.head_ref == 'next') steps: - - uses: actions/checkout@v4 + - uses: actions/checkout@v6 - name: Check release environment run: | diff --git a/.stats.yml b/.stats.yml index d321100e..0d591538 100644 --- a/.stats.yml +++ b/.stats.yml @@ -1,4 +1,4 @@ -configured_endpoints: 189 -openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/digitalocean%2Fgradient-0778b2e9d56c826f92ee69ef081d8d73fd94c139b85e11becaa88bf1cbe95fb9.yml -openapi_spec_hash: 49daca0dd735cad7200ca1c741a5dd43 -config_hash: fad48c8ac796b240fe3b90181586d1a4 +configured_endpoints: 193 +openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/digitalocean%2Fgradient-2344b44246a44d39ad5b74d3077bd2958745aad67feb15970756532fa0b3f9d6.yml +openapi_spec_hash: a1913979235ce152a8dc380fabe5362e +config_hash: 6c9a04f3cc5dd88e1e4f0ae42d98ba9a diff --git a/api.md b/api.md index caf241a4..45e4eaeb 100644 --- a/api.md +++ b/api.md @@ -10,6 +10,8 @@ from gradient.types import ( ChatCompletionChunk, ChatCompletionTokenLogprob, CompletionUsage, + CreateResponseResponse, + CreateResponseStreamResponse, DiskInfo, Droplet, DropletNextBackupWindow, @@ -404,6 +406,12 @@ Methods: - client.images.generate(\*\*params) -> ImageGenerateResponse +# Responses + +Methods: + +- client.responses.create(\*\*params) -> CreateResponseResponse + # GPUDroplets Types: @@ -856,6 +864,7 @@ from gradient.types.knowledge_bases import ( APIWebCrawlerDataSource, AwsDataSource, DataSourceCreateResponse, + DataSourceUpdateResponse, DataSourceListResponse, DataSourceDeleteResponse, DataSourceCreatePresignedURLsResponse, @@ -865,6 +874,7 @@ from gradient.types.knowledge_bases import ( Methods: - client.knowledge_bases.data_sources.create(path_knowledge_base_uuid, \*\*params) -> DataSourceCreateResponse +- client.knowledge_bases.data_sources.update(path_data_source_uuid, \*, path_knowledge_base_uuid, \*\*params) -> DataSourceUpdateResponse - client.knowledge_bases.data_sources.list(knowledge_base_uuid, \*\*params) -> DataSourceListResponse - client.knowledge_bases.data_sources.delete(data_source_uuid, \*, knowledge_base_uuid) -> DataSourceDeleteResponse - client.knowledge_bases.data_sources.create_presigned_urls(\*\*params) -> DataSourceCreatePresignedURLsResponse @@ -1039,3 +1049,29 @@ from gradient.types import RetrieveDocumentsResponse Methods: - client.retrieve.documents(knowledge_base_id, \*\*params) -> RetrieveDocumentsResponse + +# Apps + +## JobInvocations + +Types: + +```python +from gradient.types.apps import JobInvocationCancelResponse +``` + +Methods: + +- client.apps.job_invocations.cancel(job_invocation_id, \*, app_id, \*\*params) -> JobInvocationCancelResponse + +# Billing + +Types: + +```python +from gradient.types import BillingListInsightsResponse +``` + +Methods: + +- client.billing.list_insights(end_date, \*, account_urn, start_date, \*\*params) -> BillingListInsightsResponse diff --git a/requirements-dev.lock b/requirements-dev.lock index 63b7bd64..667e0dff 100644 --- a/requirements-dev.lock +++ b/requirements-dev.lock @@ -12,14 +12,14 @@ -e file:. aiohappyeyeballs==2.6.1 # via aiohttp -aiohttp==3.13.2 +aiohttp==3.13.3 # via gradient # via httpx-aiohttp aiosignal==1.4.0 # via aiohttp annotated-types==0.7.0 # via pydantic -anyio==4.12.0 +anyio==4.12.1 # via gradient # via httpx argcomplete==3.6.3 @@ -31,7 +31,7 @@ attrs==25.4.0 # via nox backports-asyncio-runner==1.2.0 # via pytest-asyncio -certifi==2025.11.12 +certifi==2026.1.4 # via httpcore # via httpx colorlog==6.10.1 @@ -61,7 +61,7 @@ httpx==0.28.1 # via gradient # via httpx-aiohttp # via respx -httpx-aiohttp==0.1.9 +httpx-aiohttp==0.1.12 # via gradient humanize==4.13.0 # via nox @@ -69,7 +69,7 @@ idna==3.11 # via anyio # via httpx # via yarl -importlib-metadata==8.7.0 +importlib-metadata==8.7.1 iniconfig==2.1.0 # via pytest markdown-it-py==3.0.0 @@ -82,14 +82,14 @@ multidict==6.7.0 mypy==1.17.0 mypy-extensions==1.1.0 # via mypy -nodeenv==1.9.1 +nodeenv==1.10.0 # via pyright nox==2025.11.12 packaging==25.0 # via dependency-groups # via nox # via pytest -pathspec==0.12.1 +pathspec==1.0.3 # via mypy platformdirs==4.4.0 # via virtualenv @@ -115,13 +115,13 @@ python-dateutil==2.9.0.post0 # via time-machine respx==0.22.0 rich==14.2.0 -ruff==0.14.7 +ruff==0.14.13 six==1.17.0 # via python-dateutil sniffio==1.3.1 # via gradient time-machine==2.19.0 -tomli==2.3.0 +tomli==2.4.0 # via dependency-groups # via mypy # via nox @@ -141,7 +141,7 @@ typing-extensions==4.15.0 # via virtualenv typing-inspection==0.4.2 # via pydantic -virtualenv==20.35.4 +virtualenv==20.36.1 # via nox yarl==1.22.0 # via aiohttp diff --git a/requirements.lock b/requirements.lock index b2623a7b..b48c65ea 100644 --- a/requirements.lock +++ b/requirements.lock @@ -12,21 +12,21 @@ -e file:. aiohappyeyeballs==2.6.1 # via aiohttp -aiohttp==3.13.2 +aiohttp==3.13.3 # via gradient # via httpx-aiohttp aiosignal==1.4.0 # via aiohttp annotated-types==0.7.0 # via pydantic -anyio==4.12.0 +anyio==4.12.1 # via gradient # via httpx async-timeout==5.0.1 # via aiohttp attrs==25.4.0 # via aiohttp -certifi==2025.11.12 +certifi==2026.1.4 # via httpcore # via httpx distro==1.9.0 @@ -43,7 +43,7 @@ httpcore==1.0.9 httpx==0.28.1 # via gradient # via httpx-aiohttp -httpx-aiohttp==0.1.9 +httpx-aiohttp==0.1.12 # via gradient idna==3.11 # via anyio diff --git a/src/gradient/_base_client.py b/src/gradient/_base_client.py index f038b215..e2135952 100644 --- a/src/gradient/_base_client.py +++ b/src/gradient/_base_client.py @@ -9,6 +9,7 @@ import inspect import logging import platform +import warnings import email.utils from types import TracebackType from random import random @@ -51,9 +52,11 @@ ResponseT, AnyMapping, PostParser, + BinaryTypes, RequestFiles, HttpxSendArgs, RequestOptions, + AsyncBinaryTypes, HttpxRequestFiles, ModelBuilderProtocol, not_given, @@ -83,6 +86,7 @@ APIConnectionError, APIResponseValidationError, ) +from ._utils._json import openapi_dumps log: logging.Logger = logging.getLogger(__name__) @@ -481,8 +485,19 @@ def _build_request( retries_taken: int = 0, ) -> httpx.Request: if log.isEnabledFor(logging.DEBUG): - log.debug("Request options: %s", model_dump(options, exclude_unset=True)) - + log.debug( + "Request options: %s", + model_dump( + options, + exclude_unset=True, + # Pydantic v1 can't dump every type we support in content, so we exclude it for now. + exclude={ + "content", + } + if PYDANTIC_V1 + else {}, + ), + ) kwargs: dict[str, Any] = {} json_data = options.json_data @@ -536,10 +551,18 @@ def _build_request( is_body_allowed = options.method.lower() != "get" if is_body_allowed: - if isinstance(json_data, bytes): + if options.content is not None and json_data is not None: + raise TypeError("Passing both `content` and `json_data` is not supported") + if options.content is not None and files is not None: + raise TypeError("Passing both `content` and `files` is not supported") + if options.content is not None: + kwargs["content"] = options.content + elif isinstance(json_data, bytes): kwargs["content"] = json_data - else: - kwargs["json"] = json_data if is_given(json_data) else None + elif not files: + # Don't set content when JSON is sent as multipart/form-data, + # since httpx's content param overrides other body arguments + kwargs["content"] = openapi_dumps(json_data) if is_given(json_data) and json_data is not None else None kwargs["files"] = files else: headers.pop("Content-Type", None) @@ -1210,6 +1233,7 @@ def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, options: RequestOptions = {}, files: RequestFiles | None = None, stream: Literal[False] = False, @@ -1222,6 +1246,7 @@ def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, options: RequestOptions = {}, files: RequestFiles | None = None, stream: Literal[True], @@ -1235,6 +1260,7 @@ def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, options: RequestOptions = {}, files: RequestFiles | None = None, stream: bool, @@ -1247,17 +1273,25 @@ def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, options: RequestOptions = {}, files: RequestFiles | None = None, stream: bool = False, stream_cls: type[_StreamT] | None = None, ) -> ResponseT | _StreamT: + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if files is not None and content is not None: + raise TypeError("Passing both `files` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) opts = FinalRequestOptions.construct( - method="post", - url=path, - json_data=body, - files=to_httpx_files(files), - **options, + method="post", url=path, json_data=body, content=content, files=to_httpx_files(files), **options ) return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)) @@ -1267,11 +1301,23 @@ def patch( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, ) -> ResponseT: + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if files is not None and content is not None: + raise TypeError("Passing both `files` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) opts = FinalRequestOptions.construct( - method="patch", url=path, json_data=body, files=to_httpx_files(files), **options + method="patch", url=path, json_data=body, content=content, files=to_httpx_files(files), **options ) return self.request(cast_to, opts) @@ -1281,15 +1327,23 @@ def put( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, ) -> ResponseT: + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if files is not None and content is not None: + raise TypeError("Passing both `files` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) opts = FinalRequestOptions.construct( - method="put", - url=path, - json_data=body, - files=to_httpx_files(files), - **options, + method="put", url=path, json_data=body, content=content, files=to_httpx_files(files), **options ) return self.request(cast_to, opts) @@ -1299,9 +1353,19 @@ def delete( *, cast_to: Type[ResponseT], body: Body | None = None, + content: BinaryTypes | None = None, options: RequestOptions = {}, ) -> ResponseT: - opts = FinalRequestOptions.construct(method="delete", url=path, json_data=body, **options) + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) + opts = FinalRequestOptions.construct(method="delete", url=path, json_data=body, content=content, **options) return self.request(cast_to, opts) def get_api_list( @@ -1751,6 +1815,7 @@ async def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, stream: Literal[False] = False, @@ -1763,6 +1828,7 @@ async def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, stream: Literal[True], @@ -1776,6 +1842,7 @@ async def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, stream: bool, @@ -1788,17 +1855,25 @@ async def post( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, stream: bool = False, stream_cls: type[_AsyncStreamT] | None = None, ) -> ResponseT | _AsyncStreamT: + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if files is not None and content is not None: + raise TypeError("Passing both `files` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) opts = FinalRequestOptions.construct( - method="post", - url=path, - json_data=body, - files=await async_to_httpx_files(files), - **options, + method="post", url=path, json_data=body, content=content, files=await async_to_httpx_files(files), **options ) return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls) @@ -1808,11 +1883,28 @@ async def patch( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, ) -> ResponseT: + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if files is not None and content is not None: + raise TypeError("Passing both `files` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) opts = FinalRequestOptions.construct( - method="patch", url=path, json_data=body, files=await async_to_httpx_files(files), **options + method="patch", + url=path, + json_data=body, + content=content, + files=await async_to_httpx_files(files), + **options, ) return await self.request(cast_to, opts) @@ -1822,15 +1914,23 @@ async def put( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, files: RequestFiles | None = None, options: RequestOptions = {}, ) -> ResponseT: + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if files is not None and content is not None: + raise TypeError("Passing both `files` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) opts = FinalRequestOptions.construct( - method="put", - url=path, - json_data=body, - files=await async_to_httpx_files(files), - **options, + method="put", url=path, json_data=body, content=content, files=await async_to_httpx_files(files), **options ) return await self.request(cast_to, opts) @@ -1840,9 +1940,19 @@ async def delete( *, cast_to: Type[ResponseT], body: Body | None = None, + content: AsyncBinaryTypes | None = None, options: RequestOptions = {}, ) -> ResponseT: - opts = FinalRequestOptions.construct(method="delete", url=path, json_data=body, **options) + if body is not None and content is not None: + raise TypeError("Passing both `body` and `content` is not supported") + if isinstance(body, bytes): + warnings.warn( + "Passing raw bytes as `body` is deprecated and will be removed in a future version. " + "Please pass raw bytes via the `content` parameter instead.", + DeprecationWarning, + stacklevel=2, + ) + opts = FinalRequestOptions.construct(method="delete", url=path, json_data=body, content=content, **options) return await self.request(cast_to, opts) def get_api_list( diff --git a/src/gradient/_client.py b/src/gradient/_client.py index fc58b291..18ef0c7a 100644 --- a/src/gradient/_client.py +++ b/src/gradient/_client.py @@ -34,26 +34,28 @@ if TYPE_CHECKING: from .resources import ( nfs, + apps, chat, agents, images, models, + billing, regions, retrieve, databases, inference, + responses, gpu_droplets, knowledge_bases, ) from .resources.images import ImagesResource, AsyncImagesResource + from .resources.billing import BillingResource, AsyncBillingResource from .resources.nfs.nfs import NfsResource, AsyncNfsResource from .resources.regions import RegionsResource, AsyncRegionsResource from .resources.retrieve import RetrieveResource, AsyncRetrieveResource + from .resources.apps.apps import AppsResource, AsyncAppsResource from .resources.chat.chat import ChatResource, AsyncChatResource - from .resources.gpu_droplets import ( - GPUDropletsResource, - AsyncGPUDropletsResource, - ) + from .resources.responses import ResponsesResource, AsyncResponsesResource from .resources.agents.agents import AgentsResource, AsyncAgentsResource from .resources.models.models import ModelsResource, AsyncModelsResource from .resources.databases.databases import DatabasesResource, AsyncDatabasesResource @@ -201,6 +203,12 @@ def images(self) -> ImagesResource: return ImagesResource(self) + @cached_property + def responses(self) -> ResponsesResource: + from .resources.responses import ResponsesResource + + return ResponsesResource(self) + @cached_property def gpu_droplets(self) -> GPUDropletsResource: from .resources.gpu_droplets import GPUDropletsResource @@ -249,6 +257,18 @@ def retrieve(self) -> RetrieveResource: return RetrieveResource(self) + @cached_property + def apps(self) -> AppsResource: + from .resources.apps import AppsResource + + return AppsResource(self) + + @cached_property + def billing(self) -> BillingResource: + from .resources.billing import BillingResource + + return BillingResource(self) + @cached_property def with_raw_response(self) -> GradientWithRawResponse: return GradientWithRawResponse(self) @@ -555,6 +575,12 @@ def images(self) -> AsyncImagesResource: return AsyncImagesResource(self) + @cached_property + def responses(self) -> AsyncResponsesResource: + from .resources.responses import AsyncResponsesResource + + return AsyncResponsesResource(self) + @cached_property def gpu_droplets(self) -> AsyncGPUDropletsResource: from .resources.gpu_droplets import AsyncGPUDropletsResource @@ -603,6 +629,18 @@ def retrieve(self) -> AsyncRetrieveResource: return AsyncRetrieveResource(self) + @cached_property + def apps(self) -> AsyncAppsResource: + from .resources.apps import AsyncAppsResource + + return AsyncAppsResource(self) + + @cached_property + def billing(self) -> AsyncBillingResource: + from .resources.billing import AsyncBillingResource + + return AsyncBillingResource(self) + @cached_property def with_raw_response(self) -> AsyncGradientWithRawResponse: return AsyncGradientWithRawResponse(self) @@ -807,6 +845,12 @@ def images(self) -> images.ImagesResourceWithRawResponse: return ImagesResourceWithRawResponse(self._client.images) + @cached_property + def responses(self) -> responses.ResponsesResourceWithRawResponse: + from .resources.responses import ResponsesResourceWithRawResponse + + return ResponsesResourceWithRawResponse(self._client.responses) + @cached_property def gpu_droplets(self) -> gpu_droplets.GPUDropletsResourceWithRawResponse: from .resources.gpu_droplets import GPUDropletsResourceWithRawResponse @@ -855,6 +899,18 @@ def retrieve(self) -> retrieve.RetrieveResourceWithRawResponse: return RetrieveResourceWithRawResponse(self._client.retrieve) + @cached_property + def apps(self) -> apps.AppsResourceWithRawResponse: + from .resources.apps import AppsResourceWithRawResponse + + return AppsResourceWithRawResponse(self._client.apps) + + @cached_property + def billing(self) -> billing.BillingResourceWithRawResponse: + from .resources.billing import BillingResourceWithRawResponse + + return BillingResourceWithRawResponse(self._client.billing) + class AsyncGradientWithRawResponse: _client: AsyncGradient @@ -880,6 +936,12 @@ def images(self) -> images.AsyncImagesResourceWithRawResponse: return AsyncImagesResourceWithRawResponse(self._client.images) + @cached_property + def responses(self) -> responses.AsyncResponsesResourceWithRawResponse: + from .resources.responses import AsyncResponsesResourceWithRawResponse + + return AsyncResponsesResourceWithRawResponse(self._client.responses) + @cached_property def gpu_droplets(self) -> gpu_droplets.AsyncGPUDropletsResourceWithRawResponse: from .resources.gpu_droplets import AsyncGPUDropletsResourceWithRawResponse @@ -932,6 +994,18 @@ def retrieve(self) -> retrieve.AsyncRetrieveResourceWithRawResponse: return AsyncRetrieveResourceWithRawResponse(self._client.retrieve) + @cached_property + def apps(self) -> apps.AsyncAppsResourceWithRawResponse: + from .resources.apps import AsyncAppsResourceWithRawResponse + + return AsyncAppsResourceWithRawResponse(self._client.apps) + + @cached_property + def billing(self) -> billing.AsyncBillingResourceWithRawResponse: + from .resources.billing import AsyncBillingResourceWithRawResponse + + return AsyncBillingResourceWithRawResponse(self._client.billing) + class GradientWithStreamedResponse: _client: Gradient @@ -957,6 +1031,12 @@ def images(self) -> images.ImagesResourceWithStreamingResponse: return ImagesResourceWithStreamingResponse(self._client.images) + @cached_property + def responses(self) -> responses.ResponsesResourceWithStreamingResponse: + from .resources.responses import ResponsesResourceWithStreamingResponse + + return ResponsesResourceWithStreamingResponse(self._client.responses) + @cached_property def gpu_droplets(self) -> gpu_droplets.GPUDropletsResourceWithStreamingResponse: from .resources.gpu_droplets import GPUDropletsResourceWithStreamingResponse @@ -1009,6 +1089,18 @@ def retrieve(self) -> retrieve.RetrieveResourceWithStreamingResponse: return RetrieveResourceWithStreamingResponse(self._client.retrieve) + @cached_property + def apps(self) -> apps.AppsResourceWithStreamingResponse: + from .resources.apps import AppsResourceWithStreamingResponse + + return AppsResourceWithStreamingResponse(self._client.apps) + + @cached_property + def billing(self) -> billing.BillingResourceWithStreamingResponse: + from .resources.billing import BillingResourceWithStreamingResponse + + return BillingResourceWithStreamingResponse(self._client.billing) + class AsyncGradientWithStreamedResponse: _client: AsyncGradient @@ -1034,6 +1126,12 @@ def images(self) -> images.AsyncImagesResourceWithStreamingResponse: return AsyncImagesResourceWithStreamingResponse(self._client.images) + @cached_property + def responses(self) -> responses.AsyncResponsesResourceWithStreamingResponse: + from .resources.responses import AsyncResponsesResourceWithStreamingResponse + + return AsyncResponsesResourceWithStreamingResponse(self._client.responses) + @cached_property def gpu_droplets( self, @@ -1090,6 +1188,18 @@ def retrieve(self) -> retrieve.AsyncRetrieveResourceWithStreamingResponse: return AsyncRetrieveResourceWithStreamingResponse(self._client.retrieve) + @cached_property + def apps(self) -> apps.AsyncAppsResourceWithStreamingResponse: + from .resources.apps import AsyncAppsResourceWithStreamingResponse + + return AsyncAppsResourceWithStreamingResponse(self._client.apps) + + @cached_property + def billing(self) -> billing.AsyncBillingResourceWithStreamingResponse: + from .resources.billing import AsyncBillingResourceWithStreamingResponse + + return AsyncBillingResourceWithStreamingResponse(self._client.billing) + Client = Gradient diff --git a/src/gradient/_compat.py b/src/gradient/_compat.py index bdef67f0..786ff42a 100644 --- a/src/gradient/_compat.py +++ b/src/gradient/_compat.py @@ -139,6 +139,7 @@ def model_dump( exclude_defaults: bool = False, warnings: bool = True, mode: Literal["json", "python"] = "python", + by_alias: bool | None = None, ) -> dict[str, Any]: if (not PYDANTIC_V1) or hasattr(model, "model_dump"): return model.model_dump( @@ -148,13 +149,12 @@ def model_dump( exclude_defaults=exclude_defaults, # warnings are not supported in Pydantic v1 warnings=True if PYDANTIC_V1 else warnings, + by_alias=by_alias, ) return cast( "dict[str, Any]", model.dict( # pyright: ignore[reportDeprecated, reportUnnecessaryCast] - exclude=exclude, - exclude_unset=exclude_unset, - exclude_defaults=exclude_defaults, + exclude=exclude, exclude_unset=exclude_unset, exclude_defaults=exclude_defaults, by_alias=bool(by_alias) ), ) diff --git a/src/gradient/_models.py b/src/gradient/_models.py index ca9500b2..29070e05 100644 --- a/src/gradient/_models.py +++ b/src/gradient/_models.py @@ -3,7 +3,20 @@ import os import inspect import weakref -from typing import TYPE_CHECKING, Any, Type, Union, Generic, TypeVar, Callable, Optional, cast +from typing import ( + IO, + TYPE_CHECKING, + Any, + Type, + Union, + Generic, + TypeVar, + Callable, + Iterable, + Optional, + AsyncIterable, + cast, +) from datetime import date, datetime from typing_extensions import ( List, @@ -787,6 +800,7 @@ class FinalRequestOptionsInput(TypedDict, total=False): timeout: float | Timeout | None files: HttpxRequestFiles | None idempotency_key: str + content: Union[bytes, bytearray, IO[bytes], Iterable[bytes], AsyncIterable[bytes], None] json_data: Body extra_json: AnyMapping follow_redirects: bool @@ -805,6 +819,7 @@ class FinalRequestOptions(pydantic.BaseModel): post_parser: Union[Callable[[Any], Any], NotGiven] = NotGiven() follow_redirects: Union[bool, None] = None + content: Union[bytes, bytearray, IO[bytes], Iterable[bytes], AsyncIterable[bytes], None] = None # It should be noted that we cannot use `json` here as that would override # a BaseModel method in an incompatible fashion. json_data: Union[Body, None] = None diff --git a/src/gradient/_types.py b/src/gradient/_types.py index 65831fee..338a463d 100644 --- a/src/gradient/_types.py +++ b/src/gradient/_types.py @@ -13,9 +13,11 @@ Mapping, TypeVar, Callable, + Iterable, Iterator, Optional, Sequence, + AsyncIterable, ) from typing_extensions import ( Set, @@ -56,6 +58,13 @@ else: Base64FileInput = Union[IO[bytes], PathLike] FileContent = Union[IO[bytes], bytes, PathLike] # PathLike is not subscriptable in Python 3.8. + + +# Used for sending raw binary data / streaming data in request bodies +# e.g. for file uploads without multipart encoding +BinaryTypes = Union[bytes, bytearray, IO[bytes], Iterable[bytes]] +AsyncBinaryTypes = Union[bytes, bytearray, IO[bytes], AsyncIterable[bytes]] + FileTypes = Union[ # file (or bytes) FileContent, diff --git a/src/gradient/_utils/_compat.py b/src/gradient/_utils/_compat.py index dd703233..2c70b299 100644 --- a/src/gradient/_utils/_compat.py +++ b/src/gradient/_utils/_compat.py @@ -26,7 +26,7 @@ def is_union(tp: Optional[Type[Any]]) -> bool: else: import types - return tp is Union or tp is types.UnionType + return tp is Union or tp is types.UnionType # type: ignore[comparison-overlap] def is_typeddict(tp: Type[Any]) -> bool: diff --git a/src/gradient/_utils/_json.py b/src/gradient/_utils/_json.py new file mode 100644 index 00000000..60584214 --- /dev/null +++ b/src/gradient/_utils/_json.py @@ -0,0 +1,35 @@ +import json +from typing import Any +from datetime import datetime +from typing_extensions import override + +import pydantic + +from .._compat import model_dump + + +def openapi_dumps(obj: Any) -> bytes: + """ + Serialize an object to UTF-8 encoded JSON bytes. + + Extends the standard json.dumps with support for additional types + commonly used in the SDK, such as `datetime`, `pydantic.BaseModel`, etc. + """ + return json.dumps( + obj, + cls=_CustomEncoder, + # Uses the same defaults as httpx's JSON serialization + ensure_ascii=False, + separators=(",", ":"), + allow_nan=False, + ).encode() + + +class _CustomEncoder(json.JSONEncoder): + @override + def default(self, o: Any) -> Any: + if isinstance(o, datetime): + return o.isoformat() + if isinstance(o, pydantic.BaseModel): + return model_dump(o, exclude_unset=True, mode="json", by_alias=True) + return super().default(o) diff --git a/src/gradient/resources/__init__.py b/src/gradient/resources/__init__.py index f668bb06..6f218273 100644 --- a/src/gradient/resources/__init__.py +++ b/src/gradient/resources/__init__.py @@ -8,6 +8,14 @@ NfsResourceWithStreamingResponse, AsyncNfsResourceWithStreamingResponse, ) +from .apps import ( + AppsResource, + AsyncAppsResource, + AppsResourceWithRawResponse, + AsyncAppsResourceWithRawResponse, + AppsResourceWithStreamingResponse, + AsyncAppsResourceWithStreamingResponse, +) from .chat import ( ChatResource, AsyncChatResource, @@ -40,6 +48,14 @@ ModelsResourceWithStreamingResponse, AsyncModelsResourceWithStreamingResponse, ) +from .billing import ( + BillingResource, + AsyncBillingResource, + BillingResourceWithRawResponse, + AsyncBillingResourceWithRawResponse, + BillingResourceWithStreamingResponse, + AsyncBillingResourceWithStreamingResponse, +) from .regions import ( RegionsResource, AsyncRegionsResource, @@ -72,6 +88,14 @@ InferenceResourceWithStreamingResponse, AsyncInferenceResourceWithStreamingResponse, ) +from .responses import ( + ResponsesResource, + AsyncResponsesResource, + ResponsesResourceWithRawResponse, + AsyncResponsesResourceWithRawResponse, + ResponsesResourceWithStreamingResponse, + AsyncResponsesResourceWithStreamingResponse, +) from .gpu_droplets import ( GPUDropletsResource, AsyncGPUDropletsResource, @@ -108,6 +132,12 @@ "AsyncImagesResourceWithRawResponse", "ImagesResourceWithStreamingResponse", "AsyncImagesResourceWithStreamingResponse", + "ResponsesResource", + "AsyncResponsesResource", + "ResponsesResourceWithRawResponse", + "AsyncResponsesResourceWithRawResponse", + "ResponsesResourceWithStreamingResponse", + "AsyncResponsesResourceWithStreamingResponse", "GPUDropletsResource", "AsyncGPUDropletsResource", "GPUDropletsResourceWithRawResponse", @@ -156,4 +186,16 @@ "AsyncRetrieveResourceWithRawResponse", "RetrieveResourceWithStreamingResponse", "AsyncRetrieveResourceWithStreamingResponse", + "AppsResource", + "AsyncAppsResource", + "AppsResourceWithRawResponse", + "AsyncAppsResourceWithRawResponse", + "AppsResourceWithStreamingResponse", + "AsyncAppsResourceWithStreamingResponse", + "BillingResource", + "AsyncBillingResource", + "BillingResourceWithRawResponse", + "AsyncBillingResourceWithRawResponse", + "BillingResourceWithStreamingResponse", + "AsyncBillingResourceWithStreamingResponse", ] diff --git a/src/gradient/resources/agents/evaluation_datasets.py b/src/gradient/resources/agents/evaluation_datasets.py index 0f9631ba..db9b473d 100644 --- a/src/gradient/resources/agents/evaluation_datasets.py +++ b/src/gradient/resources/agents/evaluation_datasets.py @@ -3,6 +3,7 @@ from __future__ import annotations from typing import Iterable +from typing_extensions import Literal import httpx @@ -53,6 +54,10 @@ def with_streaming_response(self) -> EvaluationDatasetsResourceWithStreamingResp def create( self, *, + dataset_type: Literal[ + "EVALUATION_DATASET_TYPE_UNKNOWN", "EVALUATION_DATASET_TYPE_ADK", "EVALUATION_DATASET_TYPE_NON_ADK" + ] + | Omit = omit, file_upload_dataset: APIFileUploadDataSourceParam | Omit = omit, name: str | Omit = omit, # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. @@ -85,6 +90,7 @@ def create( else "https://api.digitalocean.com/v2/gen-ai/evaluation_datasets", body=maybe_transform( { + "dataset_type": dataset_type, "file_upload_dataset": file_upload_dataset, "name": name, }, @@ -160,6 +166,10 @@ def with_streaming_response(self) -> AsyncEvaluationDatasetsResourceWithStreamin async def create( self, *, + dataset_type: Literal[ + "EVALUATION_DATASET_TYPE_UNKNOWN", "EVALUATION_DATASET_TYPE_ADK", "EVALUATION_DATASET_TYPE_NON_ADK" + ] + | Omit = omit, file_upload_dataset: APIFileUploadDataSourceParam | Omit = omit, name: str | Omit = omit, # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. @@ -192,6 +202,7 @@ async def create( else "https://api.digitalocean.com/v2/gen-ai/evaluation_datasets", body=await async_maybe_transform( { + "dataset_type": dataset_type, "file_upload_dataset": file_upload_dataset, "name": name, }, diff --git a/src/gradient/resources/agents/evaluation_runs.py b/src/gradient/resources/agents/evaluation_runs.py index 8506b00f..2b5745af 100644 --- a/src/gradient/resources/agents/evaluation_runs.py +++ b/src/gradient/resources/agents/evaluation_runs.py @@ -47,6 +47,7 @@ def with_streaming_response(self) -> EvaluationRunsResourceWithStreamingResponse def create( self, *, + agent_deployment_names: SequenceNotStr[str] | Omit = omit, agent_uuids: SequenceNotStr[str] | Omit = omit, run_name: str | Omit = omit, test_case_uuid: str | Omit = omit, @@ -62,7 +63,9 @@ def create( `/v2/gen-ai/evaluation_runs`. Args: - agent_uuids: Agent UUIDs to run the test case against. + agent_deployment_names: Agent deployment names to run the test case against (ADK agent workspaces). + + agent_uuids: Agent UUIDs to run the test case against (legacy agents). run_name: The name of the run. @@ -82,6 +85,7 @@ def create( else "https://api.digitalocean.com/v2/gen-ai/evaluation_runs", body=maybe_transform( { + "agent_deployment_names": agent_deployment_names, "agent_uuids": agent_uuids, "run_name": run_name, "test_case_uuid": test_case_uuid, @@ -249,6 +253,7 @@ def with_streaming_response(self) -> AsyncEvaluationRunsResourceWithStreamingRes async def create( self, *, + agent_deployment_names: SequenceNotStr[str] | Omit = omit, agent_uuids: SequenceNotStr[str] | Omit = omit, run_name: str | Omit = omit, test_case_uuid: str | Omit = omit, @@ -264,7 +269,9 @@ async def create( `/v2/gen-ai/evaluation_runs`. Args: - agent_uuids: Agent UUIDs to run the test case against. + agent_deployment_names: Agent deployment names to run the test case against (ADK agent workspaces). + + agent_uuids: Agent UUIDs to run the test case against (legacy agents). run_name: The name of the run. @@ -284,6 +291,7 @@ async def create( else "https://api.digitalocean.com/v2/gen-ai/evaluation_runs", body=await async_maybe_transform( { + "agent_deployment_names": agent_deployment_names, "agent_uuids": agent_uuids, "run_name": run_name, "test_case_uuid": test_case_uuid, diff --git a/src/gradient/resources/agents/evaluation_test_cases.py b/src/gradient/resources/agents/evaluation_test_cases.py index d53b8c26..0e8cce03 100644 --- a/src/gradient/resources/agents/evaluation_test_cases.py +++ b/src/gradient/resources/agents/evaluation_test_cases.py @@ -56,6 +56,7 @@ def with_streaming_response(self) -> EvaluationTestCasesResourceWithStreamingRes def create( self, *, + agent_workspace_name: str | Omit = omit, dataset_uuid: str | Omit = omit, description: str | Omit = omit, metrics: SequenceNotStr[str] | Omit = omit, @@ -98,6 +99,7 @@ def create( else "https://api.digitalocean.com/v2/gen-ai/evaluation_test_cases", body=maybe_transform( { + "agent_workspace_name": agent_workspace_name, "dataset_uuid": dataset_uuid, "description": description, "metrics": metrics, @@ -318,6 +320,7 @@ def with_streaming_response(self) -> AsyncEvaluationTestCasesResourceWithStreami async def create( self, *, + agent_workspace_name: str | Omit = omit, dataset_uuid: str | Omit = omit, description: str | Omit = omit, metrics: SequenceNotStr[str] | Omit = omit, @@ -360,6 +363,7 @@ async def create( else "https://api.digitalocean.com/v2/gen-ai/evaluation_test_cases", body=await async_maybe_transform( { + "agent_workspace_name": agent_workspace_name, "dataset_uuid": dataset_uuid, "description": description, "metrics": metrics, diff --git a/src/gradient/resources/apps/__init__.py b/src/gradient/resources/apps/__init__.py new file mode 100644 index 00000000..3033a599 --- /dev/null +++ b/src/gradient/resources/apps/__init__.py @@ -0,0 +1,33 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from .apps import ( + AppsResource, + AsyncAppsResource, + AppsResourceWithRawResponse, + AsyncAppsResourceWithRawResponse, + AppsResourceWithStreamingResponse, + AsyncAppsResourceWithStreamingResponse, +) +from .job_invocations import ( + JobInvocationsResource, + AsyncJobInvocationsResource, + JobInvocationsResourceWithRawResponse, + AsyncJobInvocationsResourceWithRawResponse, + JobInvocationsResourceWithStreamingResponse, + AsyncJobInvocationsResourceWithStreamingResponse, +) + +__all__ = [ + "JobInvocationsResource", + "AsyncJobInvocationsResource", + "JobInvocationsResourceWithRawResponse", + "AsyncJobInvocationsResourceWithRawResponse", + "JobInvocationsResourceWithStreamingResponse", + "AsyncJobInvocationsResourceWithStreamingResponse", + "AppsResource", + "AsyncAppsResource", + "AppsResourceWithRawResponse", + "AsyncAppsResourceWithRawResponse", + "AppsResourceWithStreamingResponse", + "AsyncAppsResourceWithStreamingResponse", +] diff --git a/src/gradient/resources/apps/apps.py b/src/gradient/resources/apps/apps.py new file mode 100644 index 00000000..889f2406 --- /dev/null +++ b/src/gradient/resources/apps/apps.py @@ -0,0 +1,102 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from ..._compat import cached_property +from ..._resource import SyncAPIResource, AsyncAPIResource +from .job_invocations import ( + JobInvocationsResource, + AsyncJobInvocationsResource, + JobInvocationsResourceWithRawResponse, + AsyncJobInvocationsResourceWithRawResponse, + JobInvocationsResourceWithStreamingResponse, + AsyncJobInvocationsResourceWithStreamingResponse, +) + +__all__ = ["AppsResource", "AsyncAppsResource"] + + +class AppsResource(SyncAPIResource): + @cached_property + def job_invocations(self) -> JobInvocationsResource: + return JobInvocationsResource(self._client) + + @cached_property + def with_raw_response(self) -> AppsResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return AppsResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> AppsResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return AppsResourceWithStreamingResponse(self) + + +class AsyncAppsResource(AsyncAPIResource): + @cached_property + def job_invocations(self) -> AsyncJobInvocationsResource: + return AsyncJobInvocationsResource(self._client) + + @cached_property + def with_raw_response(self) -> AsyncAppsResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return AsyncAppsResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> AsyncAppsResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return AsyncAppsResourceWithStreamingResponse(self) + + +class AppsResourceWithRawResponse: + def __init__(self, apps: AppsResource) -> None: + self._apps = apps + + @cached_property + def job_invocations(self) -> JobInvocationsResourceWithRawResponse: + return JobInvocationsResourceWithRawResponse(self._apps.job_invocations) + + +class AsyncAppsResourceWithRawResponse: + def __init__(self, apps: AsyncAppsResource) -> None: + self._apps = apps + + @cached_property + def job_invocations(self) -> AsyncJobInvocationsResourceWithRawResponse: + return AsyncJobInvocationsResourceWithRawResponse(self._apps.job_invocations) + + +class AppsResourceWithStreamingResponse: + def __init__(self, apps: AppsResource) -> None: + self._apps = apps + + @cached_property + def job_invocations(self) -> JobInvocationsResourceWithStreamingResponse: + return JobInvocationsResourceWithStreamingResponse(self._apps.job_invocations) + + +class AsyncAppsResourceWithStreamingResponse: + def __init__(self, apps: AsyncAppsResource) -> None: + self._apps = apps + + @cached_property + def job_invocations(self) -> AsyncJobInvocationsResourceWithStreamingResponse: + return AsyncJobInvocationsResourceWithStreamingResponse(self._apps.job_invocations) diff --git a/src/gradient/resources/apps/job_invocations.py b/src/gradient/resources/apps/job_invocations.py new file mode 100644 index 00000000..449dd829 --- /dev/null +++ b/src/gradient/resources/apps/job_invocations.py @@ -0,0 +1,191 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +import httpx + +from ..._types import Body, Omit, Query, Headers, NotGiven, omit, not_given +from ..._utils import maybe_transform, async_maybe_transform +from ..._compat import cached_property +from ..._resource import SyncAPIResource, AsyncAPIResource +from ..._response import ( + to_raw_response_wrapper, + to_streamed_response_wrapper, + async_to_raw_response_wrapper, + async_to_streamed_response_wrapper, +) +from ...types.apps import job_invocation_cancel_params +from ..._base_client import make_request_options +from ...types.apps.job_invocation_cancel_response import JobInvocationCancelResponse + +__all__ = ["JobInvocationsResource", "AsyncJobInvocationsResource"] + + +class JobInvocationsResource(SyncAPIResource): + @cached_property + def with_raw_response(self) -> JobInvocationsResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return JobInvocationsResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> JobInvocationsResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return JobInvocationsResourceWithStreamingResponse(self) + + def cancel( + self, + job_invocation_id: str, + *, + app_id: str, + job_name: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> JobInvocationCancelResponse: + """ + Cancel a specific job invocation for an app. + + Args: + job_name: The job name to list job invocations for. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + if not app_id: + raise ValueError(f"Expected a non-empty value for `app_id` but received {app_id!r}") + if not job_invocation_id: + raise ValueError(f"Expected a non-empty value for `job_invocation_id` but received {job_invocation_id!r}") + return self._post( + f"/v2/apps/{app_id}/job-invocations/{job_invocation_id}/cancel" + if self._client._base_url_overridden + else f"https://api.digitalocean.com/v2/apps/{app_id}/job-invocations/{job_invocation_id}/cancel", + options=make_request_options( + extra_headers=extra_headers, + extra_query=extra_query, + extra_body=extra_body, + timeout=timeout, + query=maybe_transform({"job_name": job_name}, job_invocation_cancel_params.JobInvocationCancelParams), + ), + cast_to=JobInvocationCancelResponse, + ) + + +class AsyncJobInvocationsResource(AsyncAPIResource): + @cached_property + def with_raw_response(self) -> AsyncJobInvocationsResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return AsyncJobInvocationsResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> AsyncJobInvocationsResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return AsyncJobInvocationsResourceWithStreamingResponse(self) + + async def cancel( + self, + job_invocation_id: str, + *, + app_id: str, + job_name: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> JobInvocationCancelResponse: + """ + Cancel a specific job invocation for an app. + + Args: + job_name: The job name to list job invocations for. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + if not app_id: + raise ValueError(f"Expected a non-empty value for `app_id` but received {app_id!r}") + if not job_invocation_id: + raise ValueError(f"Expected a non-empty value for `job_invocation_id` but received {job_invocation_id!r}") + return await self._post( + f"/v2/apps/{app_id}/job-invocations/{job_invocation_id}/cancel" + if self._client._base_url_overridden + else f"https://api.digitalocean.com/v2/apps/{app_id}/job-invocations/{job_invocation_id}/cancel", + options=make_request_options( + extra_headers=extra_headers, + extra_query=extra_query, + extra_body=extra_body, + timeout=timeout, + query=await async_maybe_transform( + {"job_name": job_name}, job_invocation_cancel_params.JobInvocationCancelParams + ), + ), + cast_to=JobInvocationCancelResponse, + ) + + +class JobInvocationsResourceWithRawResponse: + def __init__(self, job_invocations: JobInvocationsResource) -> None: + self._job_invocations = job_invocations + + self.cancel = to_raw_response_wrapper( + job_invocations.cancel, + ) + + +class AsyncJobInvocationsResourceWithRawResponse: + def __init__(self, job_invocations: AsyncJobInvocationsResource) -> None: + self._job_invocations = job_invocations + + self.cancel = async_to_raw_response_wrapper( + job_invocations.cancel, + ) + + +class JobInvocationsResourceWithStreamingResponse: + def __init__(self, job_invocations: JobInvocationsResource) -> None: + self._job_invocations = job_invocations + + self.cancel = to_streamed_response_wrapper( + job_invocations.cancel, + ) + + +class AsyncJobInvocationsResourceWithStreamingResponse: + def __init__(self, job_invocations: AsyncJobInvocationsResource) -> None: + self._job_invocations = job_invocations + + self.cancel = async_to_streamed_response_wrapper( + job_invocations.cancel, + ) diff --git a/src/gradient/resources/billing.py b/src/gradient/resources/billing.py new file mode 100644 index 00000000..f9f42aad --- /dev/null +++ b/src/gradient/resources/billing.py @@ -0,0 +1,226 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from typing import Union +from datetime import date + +import httpx + +from ..types import billing_list_insights_params +from .._types import Body, Omit, Query, Headers, NotGiven, omit, not_given +from .._utils import maybe_transform, async_maybe_transform +from .._compat import cached_property +from .._resource import SyncAPIResource, AsyncAPIResource +from .._response import ( + to_raw_response_wrapper, + to_streamed_response_wrapper, + async_to_raw_response_wrapper, + async_to_streamed_response_wrapper, +) +from .._base_client import make_request_options +from ..types.billing_list_insights_response import BillingListInsightsResponse + +__all__ = ["BillingResource", "AsyncBillingResource"] + + +class BillingResource(SyncAPIResource): + @cached_property + def with_raw_response(self) -> BillingResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return BillingResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> BillingResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return BillingResourceWithStreamingResponse(self) + + def list_insights( + self, + end_date: Union[str, date], + *, + account_urn: str, + start_date: Union[str, date], + page: int | Omit = omit, + per_page: int | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> BillingListInsightsResponse: + """ + This endpoint returns day-over-day changes in billing resource usage based on + nightly invoice items, including total amount, region, SKU, and description for + a specified date range. It is important to note that the daily resource usage + may not reflect month-end billing totals when totaled for a given month as + nightly invoice item estimates do not necessarily encompass all invoicing + factors for the entire month. + + Args: + page: Which 'page' of paginated results to return. + + per_page: Number of items returned per page + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + if not account_urn: + raise ValueError(f"Expected a non-empty value for `account_urn` but received {account_urn!r}") + if not start_date: + raise ValueError(f"Expected a non-empty value for `start_date` but received {start_date!r}") + if not end_date: + raise ValueError(f"Expected a non-empty value for `end_date` but received {end_date!r}") + return self._get( + f"/v2/billing/{account_urn}/insights/{start_date}/{end_date}" + if self._client._base_url_overridden + else f"https://api.digitalocean.com/v2/billing/{account_urn}/insights/{start_date}/{end_date}", + options=make_request_options( + extra_headers=extra_headers, + extra_query=extra_query, + extra_body=extra_body, + timeout=timeout, + query=maybe_transform( + { + "page": page, + "per_page": per_page, + }, + billing_list_insights_params.BillingListInsightsParams, + ), + ), + cast_to=BillingListInsightsResponse, + ) + + +class AsyncBillingResource(AsyncAPIResource): + @cached_property + def with_raw_response(self) -> AsyncBillingResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return AsyncBillingResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> AsyncBillingResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return AsyncBillingResourceWithStreamingResponse(self) + + async def list_insights( + self, + end_date: Union[str, date], + *, + account_urn: str, + start_date: Union[str, date], + page: int | Omit = omit, + per_page: int | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> BillingListInsightsResponse: + """ + This endpoint returns day-over-day changes in billing resource usage based on + nightly invoice items, including total amount, region, SKU, and description for + a specified date range. It is important to note that the daily resource usage + may not reflect month-end billing totals when totaled for a given month as + nightly invoice item estimates do not necessarily encompass all invoicing + factors for the entire month. + + Args: + page: Which 'page' of paginated results to return. + + per_page: Number of items returned per page + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + if not account_urn: + raise ValueError(f"Expected a non-empty value for `account_urn` but received {account_urn!r}") + if not start_date: + raise ValueError(f"Expected a non-empty value for `start_date` but received {start_date!r}") + if not end_date: + raise ValueError(f"Expected a non-empty value for `end_date` but received {end_date!r}") + return await self._get( + f"/v2/billing/{account_urn}/insights/{start_date}/{end_date}" + if self._client._base_url_overridden + else f"https://api.digitalocean.com/v2/billing/{account_urn}/insights/{start_date}/{end_date}", + options=make_request_options( + extra_headers=extra_headers, + extra_query=extra_query, + extra_body=extra_body, + timeout=timeout, + query=await async_maybe_transform( + { + "page": page, + "per_page": per_page, + }, + billing_list_insights_params.BillingListInsightsParams, + ), + ), + cast_to=BillingListInsightsResponse, + ) + + +class BillingResourceWithRawResponse: + def __init__(self, billing: BillingResource) -> None: + self._billing = billing + + self.list_insights = to_raw_response_wrapper( + billing.list_insights, + ) + + +class AsyncBillingResourceWithRawResponse: + def __init__(self, billing: AsyncBillingResource) -> None: + self._billing = billing + + self.list_insights = async_to_raw_response_wrapper( + billing.list_insights, + ) + + +class BillingResourceWithStreamingResponse: + def __init__(self, billing: BillingResource) -> None: + self._billing = billing + + self.list_insights = to_streamed_response_wrapper( + billing.list_insights, + ) + + +class AsyncBillingResourceWithStreamingResponse: + def __init__(self, billing: AsyncBillingResource) -> None: + self._billing = billing + + self.list_insights = async_to_streamed_response_wrapper( + billing.list_insights, + ) diff --git a/src/gradient/resources/knowledge_bases/data_sources.py b/src/gradient/resources/knowledge_bases/data_sources.py index a00d93f5..6c339108 100644 --- a/src/gradient/resources/knowledge_bases/data_sources.py +++ b/src/gradient/resources/knowledge_bases/data_sources.py @@ -3,6 +3,7 @@ from __future__ import annotations from typing import Iterable +from typing_extensions import Literal import httpx @@ -20,12 +21,14 @@ from ...types.knowledge_bases import ( data_source_list_params, data_source_create_params, + data_source_update_params, data_source_create_presigned_urls_params, ) from ...types.knowledge_bases.aws_data_source_param import AwsDataSourceParam from ...types.knowledge_bases.data_source_list_response import DataSourceListResponse from ...types.knowledge_bases.data_source_create_response import DataSourceCreateResponse from ...types.knowledge_bases.data_source_delete_response import DataSourceDeleteResponse +from ...types.knowledge_bases.data_source_update_response import DataSourceUpdateResponse from ...types.knowledge_bases.api_spaces_data_source_param import APISpacesDataSourceParam from ...types.knowledge_bases.api_web_crawler_data_source_param import APIWebCrawlerDataSourceParam from ...types.knowledge_bases.data_source_create_presigned_urls_response import DataSourceCreatePresignedURLsResponse @@ -58,6 +61,15 @@ def create( path_knowledge_base_uuid: str, *, aws_data_source: AwsDataSourceParam | Omit = omit, + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + | Omit = omit, + chunking_options: data_source_create_params.ChunkingOptions | Omit = omit, body_knowledge_base_uuid: str | Omit = omit, spaces_data_source: APISpacesDataSourceParam | Omit = omit, web_crawler_data_source: APIWebCrawlerDataSourceParam | Omit = omit, @@ -75,6 +87,16 @@ def create( Args: aws_data_source: AWS S3 Data Source + chunking_algorithm: The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + + chunking_options: Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + body_knowledge_base_uuid: Knowledge base id spaces_data_source: Spaces Bucket Data Source @@ -100,6 +122,8 @@ def create( body=maybe_transform( { "aws_data_source": aws_data_source, + "chunking_algorithm": chunking_algorithm, + "chunking_options": chunking_options, "body_knowledge_base_uuid": body_knowledge_base_uuid, "spaces_data_source": spaces_data_source, "web_crawler_data_source": web_crawler_data_source, @@ -112,6 +136,84 @@ def create( cast_to=DataSourceCreateResponse, ) + def update( + self, + path_data_source_uuid: str, + *, + path_knowledge_base_uuid: str, + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + | Omit = omit, + chunking_options: data_source_update_params.ChunkingOptions | Omit = omit, + body_data_source_uuid: str | Omit = omit, + body_knowledge_base_uuid: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> DataSourceUpdateResponse: + """To update a data source (e.g. + + chunking options), send a PUT request to + `/v2/gen-ai/knowledge_bases/{knowledge_base_uuid}/data_sources/{data_source_uuid}`. + + Args: + chunking_algorithm: The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + + chunking_options: Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + + body_data_source_uuid: Data Source ID (Path Parameter) + + body_knowledge_base_uuid: Knowledge Base ID (Path Parameter) + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + if not path_knowledge_base_uuid: + raise ValueError( + f"Expected a non-empty value for `path_knowledge_base_uuid` but received {path_knowledge_base_uuid!r}" + ) + if not path_data_source_uuid: + raise ValueError( + f"Expected a non-empty value for `path_data_source_uuid` but received {path_data_source_uuid!r}" + ) + return self._put( + f"/v2/gen-ai/knowledge_bases/{path_knowledge_base_uuid}/data_sources/{path_data_source_uuid}" + if self._client._base_url_overridden + else f"https://api.digitalocean.com/v2/gen-ai/knowledge_bases/{path_knowledge_base_uuid}/data_sources/{path_data_source_uuid}", + body=maybe_transform( + { + "chunking_algorithm": chunking_algorithm, + "chunking_options": chunking_options, + "body_data_source_uuid": body_data_source_uuid, + "body_knowledge_base_uuid": body_knowledge_base_uuid, + }, + data_source_update_params.DataSourceUpdateParams, + ), + options=make_request_options( + extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout + ), + cast_to=DataSourceUpdateResponse, + ) + def list( self, knowledge_base_uuid: str, @@ -272,6 +374,15 @@ async def create( path_knowledge_base_uuid: str, *, aws_data_source: AwsDataSourceParam | Omit = omit, + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + | Omit = omit, + chunking_options: data_source_create_params.ChunkingOptions | Omit = omit, body_knowledge_base_uuid: str | Omit = omit, spaces_data_source: APISpacesDataSourceParam | Omit = omit, web_crawler_data_source: APIWebCrawlerDataSourceParam | Omit = omit, @@ -289,6 +400,16 @@ async def create( Args: aws_data_source: AWS S3 Data Source + chunking_algorithm: The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + + chunking_options: Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + body_knowledge_base_uuid: Knowledge base id spaces_data_source: Spaces Bucket Data Source @@ -314,6 +435,8 @@ async def create( body=await async_maybe_transform( { "aws_data_source": aws_data_source, + "chunking_algorithm": chunking_algorithm, + "chunking_options": chunking_options, "body_knowledge_base_uuid": body_knowledge_base_uuid, "spaces_data_source": spaces_data_source, "web_crawler_data_source": web_crawler_data_source, @@ -326,6 +449,84 @@ async def create( cast_to=DataSourceCreateResponse, ) + async def update( + self, + path_data_source_uuid: str, + *, + path_knowledge_base_uuid: str, + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + | Omit = omit, + chunking_options: data_source_update_params.ChunkingOptions | Omit = omit, + body_data_source_uuid: str | Omit = omit, + body_knowledge_base_uuid: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> DataSourceUpdateResponse: + """To update a data source (e.g. + + chunking options), send a PUT request to + `/v2/gen-ai/knowledge_bases/{knowledge_base_uuid}/data_sources/{data_source_uuid}`. + + Args: + chunking_algorithm: The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + + chunking_options: Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + + body_data_source_uuid: Data Source ID (Path Parameter) + + body_knowledge_base_uuid: Knowledge Base ID (Path Parameter) + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + if not path_knowledge_base_uuid: + raise ValueError( + f"Expected a non-empty value for `path_knowledge_base_uuid` but received {path_knowledge_base_uuid!r}" + ) + if not path_data_source_uuid: + raise ValueError( + f"Expected a non-empty value for `path_data_source_uuid` but received {path_data_source_uuid!r}" + ) + return await self._put( + f"/v2/gen-ai/knowledge_bases/{path_knowledge_base_uuid}/data_sources/{path_data_source_uuid}" + if self._client._base_url_overridden + else f"https://api.digitalocean.com/v2/gen-ai/knowledge_bases/{path_knowledge_base_uuid}/data_sources/{path_data_source_uuid}", + body=await async_maybe_transform( + { + "chunking_algorithm": chunking_algorithm, + "chunking_options": chunking_options, + "body_data_source_uuid": body_data_source_uuid, + "body_knowledge_base_uuid": body_knowledge_base_uuid, + }, + data_source_update_params.DataSourceUpdateParams, + ), + options=make_request_options( + extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout + ), + cast_to=DataSourceUpdateResponse, + ) + async def list( self, knowledge_base_uuid: str, @@ -468,6 +669,9 @@ def __init__(self, data_sources: DataSourcesResource) -> None: self.create = to_raw_response_wrapper( data_sources.create, ) + self.update = to_raw_response_wrapper( + data_sources.update, + ) self.list = to_raw_response_wrapper( data_sources.list, ) @@ -486,6 +690,9 @@ def __init__(self, data_sources: AsyncDataSourcesResource) -> None: self.create = async_to_raw_response_wrapper( data_sources.create, ) + self.update = async_to_raw_response_wrapper( + data_sources.update, + ) self.list = async_to_raw_response_wrapper( data_sources.list, ) @@ -504,6 +711,9 @@ def __init__(self, data_sources: DataSourcesResource) -> None: self.create = to_streamed_response_wrapper( data_sources.create, ) + self.update = to_streamed_response_wrapper( + data_sources.update, + ) self.list = to_streamed_response_wrapper( data_sources.list, ) @@ -522,6 +732,9 @@ def __init__(self, data_sources: AsyncDataSourcesResource) -> None: self.create = async_to_streamed_response_wrapper( data_sources.create, ) + self.update = async_to_streamed_response_wrapper( + data_sources.update, + ) self.list = async_to_streamed_response_wrapper( data_sources.list, ) diff --git a/src/gradient/resources/responses.py b/src/gradient/resources/responses.py new file mode 100644 index 00000000..a342a78d --- /dev/null +++ b/src/gradient/resources/responses.py @@ -0,0 +1,860 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from typing import Dict, List, Union, Iterable, Optional +from typing_extensions import Literal, overload + +import httpx + +from ..types import response_create_params +from .._types import Body, Omit, Query, Headers, NotGiven, SequenceNotStr, omit, not_given +from .._utils import required_args, maybe_transform, async_maybe_transform +from .._compat import cached_property +from .._resource import SyncAPIResource, AsyncAPIResource +from .._response import ( + to_raw_response_wrapper, + to_streamed_response_wrapper, + async_to_raw_response_wrapper, + async_to_streamed_response_wrapper, +) +from .._streaming import Stream, AsyncStream +from .._base_client import make_request_options +from ..types.shared.create_response_response import CreateResponseResponse +from ..types.shared.create_response_stream_response import CreateResponseStreamResponse + +__all__ = ["ResponsesResource", "AsyncResponsesResource"] + + +class ResponsesResource(SyncAPIResource): + @cached_property + def with_raw_response(self) -> ResponsesResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return ResponsesResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> ResponsesResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return ResponsesResourceWithStreamingResponse(self) + + @overload + def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream: Optional[Literal[False]] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> CreateResponseResponse: + """Generate text responses from text prompts. + + This endpoint supports both streaming + and non-streaming responses for VLLM models only. + + Args: + input: The input text prompt or conversation history. Can be a string or an array of + message objects for conversation context. + + model: Model ID used to generate the response. Must be a VLLM model. + + instructions: System-level instructions for the model. This sets the behavior and context for + the response generation. + + max_output_tokens: Maximum number of tokens to generate in the response. If not specified, the + model will use a default value. + + max_tokens: The maximum number of tokens that can be generated in the completion. Alias for + max_output_tokens for compatibility. + + metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful + for storing additional information about the object in a structured format, and + querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + + modalities: Specifies the output types the model should generate. For text-to-text, this + should be ["text"]. + + parallel_tool_calls: Whether to enable parallel tool calls. When true, the model can make multiple + tool calls in parallel. + + stop: Up to 4 sequences where the API will stop generating further tokens. The + returned text will not contain the stop sequence. + + stream: If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + + stream_options: Options for streaming response. Only set this when you set `stream: true`. + + temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will + make the output more random, while lower values like 0.2 will make it more + focused and deterministic. We generally recommend altering this or `top_p` but + not both. + + tool_choice: Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + + tools: A list of tools the model may call. Currently, only functions are supported as a + tool. Uses Responses API format (with `name`, `description`, `parameters` at top + level). + + top_p: An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + + user: A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + ... + + @overload + def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + stream: Literal[True], + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> Stream[CreateResponseStreamResponse]: + """Generate text responses from text prompts. + + This endpoint supports both streaming + and non-streaming responses for VLLM models only. + + Args: + input: The input text prompt or conversation history. Can be a string or an array of + message objects for conversation context. + + model: Model ID used to generate the response. Must be a VLLM model. + + stream: If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + + instructions: System-level instructions for the model. This sets the behavior and context for + the response generation. + + max_output_tokens: Maximum number of tokens to generate in the response. If not specified, the + model will use a default value. + + max_tokens: The maximum number of tokens that can be generated in the completion. Alias for + max_output_tokens for compatibility. + + metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful + for storing additional information about the object in a structured format, and + querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + + modalities: Specifies the output types the model should generate. For text-to-text, this + should be ["text"]. + + parallel_tool_calls: Whether to enable parallel tool calls. When true, the model can make multiple + tool calls in parallel. + + stop: Up to 4 sequences where the API will stop generating further tokens. The + returned text will not contain the stop sequence. + + stream_options: Options for streaming response. Only set this when you set `stream: true`. + + temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will + make the output more random, while lower values like 0.2 will make it more + focused and deterministic. We generally recommend altering this or `top_p` but + not both. + + tool_choice: Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + + tools: A list of tools the model may call. Currently, only functions are supported as a + tool. Uses Responses API format (with `name`, `description`, `parameters` at top + level). + + top_p: An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + + user: A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + ... + + @overload + def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + stream: bool, + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> CreateResponseResponse | Stream[CreateResponseStreamResponse]: + """Generate text responses from text prompts. + + This endpoint supports both streaming + and non-streaming responses for VLLM models only. + + Args: + input: The input text prompt or conversation history. Can be a string or an array of + message objects for conversation context. + + model: Model ID used to generate the response. Must be a VLLM model. + + stream: If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + + instructions: System-level instructions for the model. This sets the behavior and context for + the response generation. + + max_output_tokens: Maximum number of tokens to generate in the response. If not specified, the + model will use a default value. + + max_tokens: The maximum number of tokens that can be generated in the completion. Alias for + max_output_tokens for compatibility. + + metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful + for storing additional information about the object in a structured format, and + querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + + modalities: Specifies the output types the model should generate. For text-to-text, this + should be ["text"]. + + parallel_tool_calls: Whether to enable parallel tool calls. When true, the model can make multiple + tool calls in parallel. + + stop: Up to 4 sequences where the API will stop generating further tokens. The + returned text will not contain the stop sequence. + + stream_options: Options for streaming response. Only set this when you set `stream: true`. + + temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will + make the output more random, while lower values like 0.2 will make it more + focused and deterministic. We generally recommend altering this or `top_p` but + not both. + + tool_choice: Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + + tools: A list of tools the model may call. Currently, only functions are supported as a + tool. Uses Responses API format (with `name`, `description`, `parameters` at top + level). + + top_p: An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + + user: A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + ... + + @required_args(["input", "model"], ["input", "model", "stream"]) + def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream: Optional[Literal[False]] | Literal[True] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> CreateResponseResponse | Stream[CreateResponseStreamResponse]: + return self._post( + "/responses", + body=maybe_transform( + { + "input": input, + "model": model, + "instructions": instructions, + "max_output_tokens": max_output_tokens, + "max_tokens": max_tokens, + "metadata": metadata, + "modalities": modalities, + "parallel_tool_calls": parallel_tool_calls, + "stop": stop, + "stream": stream, + "stream_options": stream_options, + "temperature": temperature, + "tool_choice": tool_choice, + "tools": tools, + "top_p": top_p, + "user": user, + }, + response_create_params.ResponseCreateParamsStreaming + if stream + else response_create_params.ResponseCreateParamsNonStreaming, + ), + options=make_request_options( + extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout + ), + cast_to=CreateResponseResponse, + stream=stream or False, + stream_cls=Stream[CreateResponseStreamResponse], + ) + + +class AsyncResponsesResource(AsyncAPIResource): + @cached_property + def with_raw_response(self) -> AsyncResponsesResourceWithRawResponse: + """ + This property can be used as a prefix for any HTTP method call to return + the raw response object instead of the parsed content. + + For more information, see https://www.github.com/digitalocean/gradient-python#accessing-raw-response-data-eg-headers + """ + return AsyncResponsesResourceWithRawResponse(self) + + @cached_property + def with_streaming_response(self) -> AsyncResponsesResourceWithStreamingResponse: + """ + An alternative to `.with_raw_response` that doesn't eagerly read the response body. + + For more information, see https://www.github.com/digitalocean/gradient-python#with_streaming_response + """ + return AsyncResponsesResourceWithStreamingResponse(self) + + @overload + async def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream: Optional[Literal[False]] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> CreateResponseResponse: + """Generate text responses from text prompts. + + This endpoint supports both streaming + and non-streaming responses for VLLM models only. + + Args: + input: The input text prompt or conversation history. Can be a string or an array of + message objects for conversation context. + + model: Model ID used to generate the response. Must be a VLLM model. + + instructions: System-level instructions for the model. This sets the behavior and context for + the response generation. + + max_output_tokens: Maximum number of tokens to generate in the response. If not specified, the + model will use a default value. + + max_tokens: The maximum number of tokens that can be generated in the completion. Alias for + max_output_tokens for compatibility. + + metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful + for storing additional information about the object in a structured format, and + querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + + modalities: Specifies the output types the model should generate. For text-to-text, this + should be ["text"]. + + parallel_tool_calls: Whether to enable parallel tool calls. When true, the model can make multiple + tool calls in parallel. + + stop: Up to 4 sequences where the API will stop generating further tokens. The + returned text will not contain the stop sequence. + + stream: If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + + stream_options: Options for streaming response. Only set this when you set `stream: true`. + + temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will + make the output more random, while lower values like 0.2 will make it more + focused and deterministic. We generally recommend altering this or `top_p` but + not both. + + tool_choice: Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + + tools: A list of tools the model may call. Currently, only functions are supported as a + tool. Uses Responses API format (with `name`, `description`, `parameters` at top + level). + + top_p: An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + + user: A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + ... + + @overload + async def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + stream: Literal[True], + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> AsyncStream[CreateResponseStreamResponse]: + """Generate text responses from text prompts. + + This endpoint supports both streaming + and non-streaming responses for VLLM models only. + + Args: + input: The input text prompt or conversation history. Can be a string or an array of + message objects for conversation context. + + model: Model ID used to generate the response. Must be a VLLM model. + + stream: If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + + instructions: System-level instructions for the model. This sets the behavior and context for + the response generation. + + max_output_tokens: Maximum number of tokens to generate in the response. If not specified, the + model will use a default value. + + max_tokens: The maximum number of tokens that can be generated in the completion. Alias for + max_output_tokens for compatibility. + + metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful + for storing additional information about the object in a structured format, and + querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + + modalities: Specifies the output types the model should generate. For text-to-text, this + should be ["text"]. + + parallel_tool_calls: Whether to enable parallel tool calls. When true, the model can make multiple + tool calls in parallel. + + stop: Up to 4 sequences where the API will stop generating further tokens. The + returned text will not contain the stop sequence. + + stream_options: Options for streaming response. Only set this when you set `stream: true`. + + temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will + make the output more random, while lower values like 0.2 will make it more + focused and deterministic. We generally recommend altering this or `top_p` but + not both. + + tool_choice: Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + + tools: A list of tools the model may call. Currently, only functions are supported as a + tool. Uses Responses API format (with `name`, `description`, `parameters` at top + level). + + top_p: An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + + user: A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + ... + + @overload + async def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + stream: bool, + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> CreateResponseResponse | AsyncStream[CreateResponseStreamResponse]: + """Generate text responses from text prompts. + + This endpoint supports both streaming + and non-streaming responses for VLLM models only. + + Args: + input: The input text prompt or conversation history. Can be a string or an array of + message objects for conversation context. + + model: Model ID used to generate the response. Must be a VLLM model. + + stream: If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + + instructions: System-level instructions for the model. This sets the behavior and context for + the response generation. + + max_output_tokens: Maximum number of tokens to generate in the response. If not specified, the + model will use a default value. + + max_tokens: The maximum number of tokens that can be generated in the completion. Alias for + max_output_tokens for compatibility. + + metadata: Set of 16 key-value pairs that can be attached to an object. This can be useful + for storing additional information about the object in a structured format, and + querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + + modalities: Specifies the output types the model should generate. For text-to-text, this + should be ["text"]. + + parallel_tool_calls: Whether to enable parallel tool calls. When true, the model can make multiple + tool calls in parallel. + + stop: Up to 4 sequences where the API will stop generating further tokens. The + returned text will not contain the stop sequence. + + stream_options: Options for streaming response. Only set this when you set `stream: true`. + + temperature: What sampling temperature to use, between 0 and 2. Higher values like 0.8 will + make the output more random, while lower values like 0.2 will make it more + focused and deterministic. We generally recommend altering this or `top_p` but + not both. + + tool_choice: Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + + tools: A list of tools the model may call. Currently, only functions are supported as a + tool. Uses Responses API format (with `name`, `description`, `parameters` at top + level). + + top_p: An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + + user: A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + + extra_headers: Send extra headers + + extra_query: Add additional query parameters to the request + + extra_body: Add additional JSON properties to the request + + timeout: Override the client-level default timeout for this request, in seconds + """ + ... + + @required_args(["input", "model"], ["input", "model", "stream"]) + async def create( + self, + *, + input: Union[str, Iterable[response_create_params.InputUnionMember1]], + model: str, + instructions: Optional[str] | Omit = omit, + max_output_tokens: Optional[int] | Omit = omit, + max_tokens: Optional[int] | Omit = omit, + metadata: Optional[Dict[str, str]] | Omit = omit, + modalities: Optional[List[Literal["text"]]] | Omit = omit, + parallel_tool_calls: Optional[bool] | Omit = omit, + stop: Union[Optional[str], SequenceNotStr[str], None] | Omit = omit, + stream: Optional[Literal[False]] | Literal[True] | Omit = omit, + stream_options: Optional[response_create_params.StreamOptions] | Omit = omit, + temperature: Optional[float] | Omit = omit, + tool_choice: response_create_params.ToolChoice | Omit = omit, + tools: Iterable[response_create_params.Tool] | Omit = omit, + top_p: Optional[float] | Omit = omit, + user: str | Omit = omit, + # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs. + # The extra values given here take precedence over values defined on the client or passed to this method. + extra_headers: Headers | None = None, + extra_query: Query | None = None, + extra_body: Body | None = None, + timeout: float | httpx.Timeout | None | NotGiven = not_given, + ) -> CreateResponseResponse | AsyncStream[CreateResponseStreamResponse]: + return await self._post( + "/responses", + body=await async_maybe_transform( + { + "input": input, + "model": model, + "instructions": instructions, + "max_output_tokens": max_output_tokens, + "max_tokens": max_tokens, + "metadata": metadata, + "modalities": modalities, + "parallel_tool_calls": parallel_tool_calls, + "stop": stop, + "stream": stream, + "stream_options": stream_options, + "temperature": temperature, + "tool_choice": tool_choice, + "tools": tools, + "top_p": top_p, + "user": user, + }, + response_create_params.ResponseCreateParamsStreaming + if stream + else response_create_params.ResponseCreateParamsNonStreaming, + ), + options=make_request_options( + extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout + ), + cast_to=CreateResponseResponse, + stream=stream or False, + stream_cls=AsyncStream[CreateResponseStreamResponse], + ) + + +class ResponsesResourceWithRawResponse: + def __init__(self, responses: ResponsesResource) -> None: + self._responses = responses + + self.create = to_raw_response_wrapper( + responses.create, + ) + + +class AsyncResponsesResourceWithRawResponse: + def __init__(self, responses: AsyncResponsesResource) -> None: + self._responses = responses + + self.create = async_to_raw_response_wrapper( + responses.create, + ) + + +class ResponsesResourceWithStreamingResponse: + def __init__(self, responses: ResponsesResource) -> None: + self._responses = responses + + self.create = to_streamed_response_wrapper( + responses.create, + ) + + +class AsyncResponsesResourceWithStreamingResponse: + def __init__(self, responses: AsyncResponsesResource) -> None: + self._responses = responses + + self.create = async_to_streamed_response_wrapper( + responses.create, + ) diff --git a/src/gradient/resources/retrieve.py b/src/gradient/resources/retrieve.py index f8335ab3..992fa37f 100644 --- a/src/gradient/resources/retrieve.py +++ b/src/gradient/resources/retrieve.py @@ -67,6 +67,12 @@ def documents( 3. Performs vector similarity search in the knowledge base 4. Returns the most relevant document chunks + The search supports hybrid search combining: + + - Vector similarity (semantic search) + - Keyword matching (BM25) + - Custom metadata filters + Args: num_results: Number of results to return @@ -162,6 +168,12 @@ async def documents( 3. Performs vector similarity search in the knowledge base 4. Returns the most relevant document chunks + The search supports hybrid search combining: + + - Vector similarity (semantic search) + - Keyword matching (BM25) + - Custom metadata filters + Args: num_results: Number of results to return diff --git a/src/gradient/types/__init__.py b/src/gradient/types/__init__.py index e927e099..fc2907b2 100644 --- a/src/gradient/types/__init__.py +++ b/src/gradient/types/__init__.py @@ -43,10 +43,12 @@ ChatCompletionChunk as ChatCompletionChunk, ImageGenStreamEvent as ImageGenStreamEvent, SubscriptionTierBase as SubscriptionTierBase, + CreateResponseResponse as CreateResponseResponse, ImageGenCompletedEvent as ImageGenCompletedEvent, DropletNextBackupWindow as DropletNextBackupWindow, ImageGenPartialImageEvent as ImageGenPartialImageEvent, ChatCompletionTokenLogprob as ChatCompletionTokenLogprob, + CreateResponseStreamResponse as CreateResponseStreamResponse, ) from .api_agent import APIAgent as APIAgent from .api_model import APIModel as APIModel @@ -77,6 +79,7 @@ from .droplet_backup_policy import DropletBackupPolicy as DropletBackupPolicy from .image_generate_params import ImageGenerateParams as ImageGenerateParams from .api_agent_api_key_info import APIAgentAPIKeyInfo as APIAgentAPIKeyInfo +from .response_create_params import ResponseCreateParams as ResponseCreateParams from .agent_retrieve_response import AgentRetrieveResponse as AgentRetrieveResponse from .api_openai_api_key_info import APIOpenAIAPIKeyInfo as APIOpenAIAPIKeyInfo from .gpu_droplet_list_params import GPUDropletListParams as GPUDropletListParams @@ -105,6 +108,7 @@ from .agent_update_status_response import ( AgentUpdateStatusResponse as AgentUpdateStatusResponse, ) +from .billing_list_insights_params import BillingListInsightsParams as BillingListInsightsParams from .knowledge_base_create_params import ( KnowledgeBaseCreateParams as KnowledgeBaseCreateParams, ) @@ -118,6 +122,9 @@ from .gpu_droplet_retrieve_response import ( GPUDropletRetrieveResponse as GPUDropletRetrieveResponse, ) +from .billing_list_insights_response import ( + BillingListInsightsResponse as BillingListInsightsResponse, +) from .knowledge_base_create_response import ( KnowledgeBaseCreateResponse as KnowledgeBaseCreateResponse, ) diff --git a/src/gradient/types/agents/api_evaluation_prompt.py b/src/gradient/types/agents/api_evaluation_prompt.py index 7471e9ae..1bb08bf1 100644 --- a/src/gradient/types/agents/api_evaluation_prompt.py +++ b/src/gradient/types/agents/api_evaluation_prompt.py @@ -1,11 +1,60 @@ # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. from typing import List, Optional +from datetime import datetime +from typing_extensions import Literal from ..._models import BaseModel from .api_evaluation_metric_result import APIEvaluationMetricResult -__all__ = ["APIEvaluationPrompt", "PromptChunk"] +__all__ = ["APIEvaluationPrompt", "EvaluationTraceSpan", "EvaluationTraceSpanRetrieverChunk", "PromptChunk"] + + +class EvaluationTraceSpanRetrieverChunk(BaseModel): + chunk_usage_pct: Optional[float] = None + """The usage percentage of the chunk.""" + + chunk_used: Optional[bool] = None + """Indicates if the chunk was used in the prompt.""" + + index_uuid: Optional[str] = None + """The index uuid (Knowledge Base) of the chunk.""" + + source_name: Optional[str] = None + """The source name for the chunk, e.g., the file name or document title.""" + + text: Optional[str] = None + """Text content of the chunk.""" + + +class EvaluationTraceSpan(BaseModel): + """Represents a span within an evaluatioin trace (e.g., LLM call, tool call, etc.)""" + + created_at: Optional[datetime] = None + """When the span was created""" + + input: Optional[object] = None + """ + Input data for the span (flexible structure - can be messages array, string, + etc.) + """ + + name: Optional[str] = None + """Name/identifier for the span""" + + output: Optional[object] = None + """Output data from the span (flexible structure - can be message, string, etc.)""" + + retriever_chunks: Optional[List[EvaluationTraceSpanRetrieverChunk]] = None + """Any retriever span chunks that were included as part of the span.""" + + span_level_metric_results: Optional[List[APIEvaluationMetricResult]] = None + """The span-level metric results.""" + + type: Optional[ + Literal["TRACE_SPAN_TYPE_UNKNOWN", "TRACE_SPAN_TYPE_LLM", "TRACE_SPAN_TYPE_RETRIEVER", "TRACE_SPAN_TYPE_TOOL"] + ] = None + """Types of spans in a trace""" class PromptChunk(BaseModel): @@ -26,6 +75,9 @@ class PromptChunk(BaseModel): class APIEvaluationPrompt(BaseModel): + evaluation_trace_spans: Optional[List[EvaluationTraceSpan]] = None + """The evaluated trace spans.""" + ground_truth: Optional[str] = None """The ground truth for the prompt.""" @@ -47,3 +99,6 @@ class APIEvaluationPrompt(BaseModel): prompt_level_metric_results: Optional[List[APIEvaluationMetricResult]] = None """The metric results for the prompt.""" + + trace_id: Optional[str] = None + """The trace id for the prompt.""" diff --git a/src/gradient/types/agents/api_evaluation_run.py b/src/gradient/types/agents/api_evaluation_run.py index 5a758898..ed4ef0a1 100644 --- a/src/gradient/types/agents/api_evaluation_run.py +++ b/src/gradient/types/agents/api_evaluation_run.py @@ -14,6 +14,9 @@ class APIEvaluationRun(BaseModel): agent_deleted: Optional[bool] = None """Whether agent is deleted""" + agent_deployment_name: Optional[str] = None + """The agent deployment name""" + agent_name: Optional[str] = None """Agent name""" diff --git a/src/gradient/types/agents/chat/completion_create_params.py b/src/gradient/types/agents/chat/completion_create_params.py index d238f8e1..797c6ea3 100644 --- a/src/gradient/types/agents/chat/completion_create_params.py +++ b/src/gradient/types/agents/chat/completion_create_params.py @@ -11,20 +11,42 @@ "CompletionCreateParamsBase", "Message", "MessageChatCompletionRequestSystemMessage", + "MessageChatCompletionRequestSystemMessageContent", + "MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestSystemMessageContentArrayOfContentPart", - "MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestDeveloperMessage", + "MessageChatCompletionRequestDeveloperMessageContent", + "MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart", - "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestUserMessage", + "MessageChatCompletionRequestUserMessageContent", + "MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestUserMessageContentArrayOfContentPart", - "MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestAssistantMessage", + "MessageChatCompletionRequestAssistantMessageContent", + "MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart", - "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestAssistantMessageToolCall", "MessageChatCompletionRequestAssistantMessageToolCallFunction", "MessageChatCompletionRequestToolMessage", + "MessageChatCompletionRequestToolMessageContent", + "MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartTextCacheControl", + "MessageChatCompletionRequestToolMessageContentArrayOfContentPart", + "MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "StreamOptions", "ToolChoice", "ToolChoiceChatCompletionNamedToolChoice", @@ -165,7 +187,46 @@ class CompletionCreateParamsBase(TypedDict, total=False): """ -class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): """Content part with type and text""" text: Required[str] @@ -174,9 +235,18 @@ class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMem type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + MessageChatCompletionRequestSystemMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestSystemMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestSystemMessageContentArrayOfContentPart], ] @@ -186,14 +256,26 @@ class MessageChatCompletionRequestSystemMessage(TypedDict, total=False): messages sent by the user. """ - content: Required[Union[str, SequenceNotStr[MessageChatCompletionRequestSystemMessageContentArrayOfContentPart]]] + content: Required[MessageChatCompletionRequestSystemMessageContent] """The contents of the system message.""" role: Required[Literal["system"]] """The role of the messages author, in this case `system`.""" -class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): """Content part with type and text""" text: Required[str] @@ -202,9 +284,45 @@ class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnion type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestDeveloperMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart], ] @@ -214,14 +332,26 @@ class MessageChatCompletionRequestDeveloperMessage(TypedDict, total=False): messages sent by the user. """ - content: Required[Union[str, SequenceNotStr[MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart]]] + content: Required[MessageChatCompletionRequestDeveloperMessageContent] """The contents of the developer message.""" role: Required[Literal["developer"]] """The role of the messages author, in this case `developer`.""" -class MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): """Content part with type and text""" text: Required[str] @@ -230,9 +360,47 @@ class MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMembe type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: ( + MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + ) + """Cache control settings for the content part.""" + MessageChatCompletionRequestUserMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestUserMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestUserMessageContentArrayOfContentPart], ] @@ -242,14 +410,26 @@ class MessageChatCompletionRequestUserMessage(TypedDict, total=False): information. """ - content: Required[Union[str, SequenceNotStr[MessageChatCompletionRequestUserMessageContentArrayOfContentPart]]] + content: Required[MessageChatCompletionRequestUserMessageContent] """The contents of the user message.""" role: Required[Literal["user"]] """The role of the messages author, in this case `user`.""" -class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): """Content part with type and text""" text: Required[str] @@ -258,9 +438,45 @@ class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnion type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestAssistantMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart], ] @@ -296,15 +512,80 @@ class MessageChatCompletionRequestAssistantMessage(TypedDict, total=False): role: Required[Literal["assistant"]] """The role of the messages author, in this case `assistant`.""" - content: Union[str, SequenceNotStr[MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart], None] + content: Optional[MessageChatCompletionRequestAssistantMessageContent] """The contents of the assistant message.""" tool_calls: Iterable[MessageChatCompletionRequestAssistantMessageToolCall] """The tool calls generated by the model, such as function calls.""" +class MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: ( + MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + ) + """Cache control settings for the content part.""" + + +MessageChatCompletionRequestToolMessageContentArrayOfContentPart: TypeAlias = Union[ + str, MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestToolMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestToolMessageContentArrayOfContentPart], +] + + class MessageChatCompletionRequestToolMessage(TypedDict, total=False): - content: Required[str] + content: Required[MessageChatCompletionRequestToolMessageContent] """The contents of the tool message.""" role: Required[Literal["tool"]] diff --git a/src/gradient/types/agents/evaluation_dataset_create_params.py b/src/gradient/types/agents/evaluation_dataset_create_params.py index c8a84c23..661a42da 100644 --- a/src/gradient/types/agents/evaluation_dataset_create_params.py +++ b/src/gradient/types/agents/evaluation_dataset_create_params.py @@ -2,7 +2,7 @@ from __future__ import annotations -from typing_extensions import TypedDict +from typing_extensions import Literal, TypedDict from ..knowledge_bases.api_file_upload_data_source_param import APIFileUploadDataSourceParam @@ -10,6 +10,10 @@ class EvaluationDatasetCreateParams(TypedDict, total=False): + dataset_type: Literal[ + "EVALUATION_DATASET_TYPE_UNKNOWN", "EVALUATION_DATASET_TYPE_ADK", "EVALUATION_DATASET_TYPE_NON_ADK" + ] + file_upload_dataset: APIFileUploadDataSourceParam """File to upload as data source for knowledge base.""" diff --git a/src/gradient/types/agents/evaluation_run_create_params.py b/src/gradient/types/agents/evaluation_run_create_params.py index 52bbee85..b5e60803 100644 --- a/src/gradient/types/agents/evaluation_run_create_params.py +++ b/src/gradient/types/agents/evaluation_run_create_params.py @@ -10,8 +10,11 @@ class EvaluationRunCreateParams(TypedDict, total=False): + agent_deployment_names: SequenceNotStr[str] + """Agent deployment names to run the test case against (ADK agent workspaces).""" + agent_uuids: SequenceNotStr[str] - """Agent UUIDs to run the test case against.""" + """Agent UUIDs to run the test case against (legacy agents).""" run_name: str """The name of the run.""" diff --git a/src/gradient/types/agents/evaluation_test_case_create_params.py b/src/gradient/types/agents/evaluation_test_case_create_params.py index af49d024..ff0666b9 100644 --- a/src/gradient/types/agents/evaluation_test_case_create_params.py +++ b/src/gradient/types/agents/evaluation_test_case_create_params.py @@ -11,6 +11,8 @@ class EvaluationTestCaseCreateParams(TypedDict, total=False): + agent_workspace_name: str + dataset_uuid: str """Dataset against which the test‑case is executed.""" diff --git a/src/gradient/types/api_agent_model.py b/src/gradient/types/api_agent_model.py index e42bb5d5..a6a36c6f 100644 --- a/src/gradient/types/api_agent_model.py +++ b/src/gradient/types/api_agent_model.py @@ -29,6 +29,15 @@ class APIAgentModel(BaseModel): is_foundational: Optional[bool] = None """True if it is a foundational model provided by do""" + kb_default_chunk_size: Optional[int] = None + """Default chunking size limit to show in UI""" + + kb_max_chunk_size: Optional[int] = None + """Maximum chunk size limit of model""" + + kb_min_chunk_size: Optional[int] = None + """Minimum chunking size token limits if model supports KNOWLEDGEBASE usecase""" + metadata: Optional[object] = None """Additional meta data""" diff --git a/src/gradient/types/api_model.py b/src/gradient/types/api_model.py index 83b1b66a..1d9752e4 100644 --- a/src/gradient/types/api_model.py +++ b/src/gradient/types/api_model.py @@ -25,6 +25,15 @@ class APIModel(BaseModel): is_foundational: Optional[bool] = None """True if it is a foundational model provided by do""" + kb_default_chunk_size: Optional[int] = None + """Default chunking size limit to show in UI""" + + kb_max_chunk_size: Optional[int] = None + """Maximum chunk size limit of model""" + + kb_min_chunk_size: Optional[int] = None + """Minimum chunking size token limits if model supports KNOWLEDGEBASE usecase""" + name: Optional[str] = None """Display name of the model""" diff --git a/src/gradient/types/apps/__init__.py b/src/gradient/types/apps/__init__.py new file mode 100644 index 00000000..b4a2e426 --- /dev/null +++ b/src/gradient/types/apps/__init__.py @@ -0,0 +1,6 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from .job_invocation_cancel_params import JobInvocationCancelParams as JobInvocationCancelParams +from .job_invocation_cancel_response import JobInvocationCancelResponse as JobInvocationCancelResponse diff --git a/src/gradient/types/apps/job_invocation_cancel_params.py b/src/gradient/types/apps/job_invocation_cancel_params.py new file mode 100644 index 00000000..b026fcb9 --- /dev/null +++ b/src/gradient/types/apps/job_invocation_cancel_params.py @@ -0,0 +1,14 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from typing_extensions import Required, TypedDict + +__all__ = ["JobInvocationCancelParams"] + + +class JobInvocationCancelParams(TypedDict, total=False): + app_id: Required[str] + + job_name: str + """The job name to list job invocations for.""" diff --git a/src/gradient/types/apps/job_invocation_cancel_response.py b/src/gradient/types/apps/job_invocation_cancel_response.py new file mode 100644 index 00000000..96b2c642 --- /dev/null +++ b/src/gradient/types/apps/job_invocation_cancel_response.py @@ -0,0 +1,77 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from typing import Optional +from datetime import datetime +from typing_extensions import Literal + +from ..._models import BaseModel + +__all__ = [ + "JobInvocationCancelResponse", + "Trigger", + "TriggerManual", + "TriggerManualUser", + "TriggerScheduled", + "TriggerScheduledSchedule", +] + + +class TriggerManualUser(BaseModel): + """The user who triggered the job""" + + email: Optional[str] = None + + full_name: Optional[str] = None + + uuid: Optional[str] = None + + +class TriggerManual(BaseModel): + """Details about the manual trigger, if applicable""" + + user: Optional[TriggerManualUser] = None + """The user who triggered the job""" + + +class TriggerScheduledSchedule(BaseModel): + cron: Optional[str] = None + """The cron expression defining the schedule""" + + time_zone: Optional[str] = None + """The time zone for the schedule""" + + +class TriggerScheduled(BaseModel): + """The schedule for the job""" + + schedule: Optional[TriggerScheduledSchedule] = None + + +class Trigger(BaseModel): + manual: Optional[TriggerManual] = None + """Details about the manual trigger, if applicable""" + + scheduled: Optional[TriggerScheduled] = None + """The schedule for the job""" + + type: Optional[Literal["MANUAL", "SCHEDULE", "UNKNOWN"]] = None + """The type of trigger that initiated the job invocation.""" + + +class JobInvocationCancelResponse(BaseModel): + id: Optional[str] = None + + completed_at: Optional[datetime] = None + + created_at: Optional[datetime] = None + + deployment_id: Optional[str] = None + + job_name: Optional[str] = None + + phase: Optional[Literal["UNKNOWN", "PENDING", "RUNNING", "SUCCEEDED", "FAILED", "CANCELED", "SKIPPED"]] = None + """The phase of the job invocation""" + + started_at: Optional[datetime] = None + + trigger: Optional[Trigger] = None diff --git a/src/gradient/types/billing_list_insights_params.py b/src/gradient/types/billing_list_insights_params.py new file mode 100644 index 00000000..6e89dd1e --- /dev/null +++ b/src/gradient/types/billing_list_insights_params.py @@ -0,0 +1,23 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from typing import Union +from datetime import date +from typing_extensions import Required, Annotated, TypedDict + +from .._utils import PropertyInfo + +__all__ = ["BillingListInsightsParams"] + + +class BillingListInsightsParams(TypedDict, total=False): + account_urn: Required[str] + + start_date: Required[Annotated[Union[str, date], PropertyInfo(format="iso8601")]] + + page: int + """Which 'page' of paginated results to return.""" + + per_page: int + """Number of items returned per page""" diff --git a/src/gradient/types/billing_list_insights_response.py b/src/gradient/types/billing_list_insights_response.py new file mode 100644 index 00000000..f7515a6b --- /dev/null +++ b/src/gradient/types/billing_list_insights_response.py @@ -0,0 +1,51 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from typing import List, Optional +from datetime import date + +from .._models import BaseModel + +__all__ = ["BillingListInsightsResponse", "DataPoint"] + + +class DataPoint(BaseModel): + description: Optional[str] = None + """Description of the billed resource or service as shown on an invoice item""" + + group_description: Optional[str] = None + """ + Optional invoice item group name of the billed resource or service, blank when + not part an invoice item group + """ + + region: Optional[str] = None + """Region where the usage occurred""" + + sku: Optional[str] = None + """Unique SKU identifier for the billed resource""" + + start_date: Optional[date] = None + """Start date of the billing data point in YYYY-MM-DD format""" + + total_amount: Optional[str] = None + """Total amount for this data point in USD""" + + usage_team_urn: Optional[str] = None + """URN of the team that incurred the usage""" + + +class BillingListInsightsResponse(BaseModel): + current_page: int + """Current page number""" + + data_points: List[DataPoint] + """ + Array of billing data points, which are day-over-day changes in billing resource + usage based on nightly invoice item estimates, for the requested period + """ + + total_items: int + """Total number of items available across all pages""" + + total_pages: int + """Total number of pages available""" diff --git a/src/gradient/types/chat/completion_create_params.py b/src/gradient/types/chat/completion_create_params.py index e889c5e8..bf5bd49d 100644 --- a/src/gradient/types/chat/completion_create_params.py +++ b/src/gradient/types/chat/completion_create_params.py @@ -11,20 +11,42 @@ "CompletionCreateParamsBase", "Message", "MessageChatCompletionRequestSystemMessage", + "MessageChatCompletionRequestSystemMessageContent", + "MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestSystemMessageContentArrayOfContentPart", - "MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestDeveloperMessage", + "MessageChatCompletionRequestDeveloperMessageContent", + "MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart", - "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestUserMessage", + "MessageChatCompletionRequestUserMessageContent", + "MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestUserMessageContentArrayOfContentPart", - "MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestAssistantMessage", + "MessageChatCompletionRequestAssistantMessageContent", + "MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart", - "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnionMember1", + "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "MessageChatCompletionRequestAssistantMessageToolCall", "MessageChatCompletionRequestAssistantMessageToolCallFunction", "MessageChatCompletionRequestToolMessage", + "MessageChatCompletionRequestToolMessageContent", + "MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartText", + "MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartTextCacheControl", + "MessageChatCompletionRequestToolMessageContentArrayOfContentPart", + "MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartText", + "MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl", "StreamOptions", "ToolChoice", "ToolChoiceChatCompletionNamedToolChoice", @@ -165,7 +187,46 @@ class CompletionCreateParamsBase(TypedDict, total=False): """ -class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): """Content part with type and text""" text: Required[str] @@ -174,9 +235,18 @@ class MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMem type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + MessageChatCompletionRequestSystemMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestSystemMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestSystemMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestSystemMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestSystemMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestSystemMessageContentArrayOfContentPart], ] @@ -186,14 +256,26 @@ class MessageChatCompletionRequestSystemMessage(TypedDict, total=False): messages sent by the user. """ - content: Required[Union[str, SequenceNotStr[MessageChatCompletionRequestSystemMessageContentArrayOfContentPart]]] + content: Required[MessageChatCompletionRequestSystemMessageContent] """The contents of the system message.""" role: Required[Literal["system"]] """The role of the messages author, in this case `system`.""" -class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): """Content part with type and text""" text: Required[str] @@ -202,9 +284,45 @@ class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnion type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestDeveloperMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestDeveloperMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart], ] @@ -214,14 +332,26 @@ class MessageChatCompletionRequestDeveloperMessage(TypedDict, total=False): messages sent by the user. """ - content: Required[Union[str, SequenceNotStr[MessageChatCompletionRequestDeveloperMessageContentArrayOfContentPart]]] + content: Required[MessageChatCompletionRequestDeveloperMessageContent] """The contents of the developer message.""" role: Required[Literal["developer"]] """The role of the messages author, in this case `developer`.""" -class MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): """Content part with type and text""" text: Required[str] @@ -230,9 +360,47 @@ class MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMembe type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: ( + MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + ) + """Cache control settings for the content part.""" + MessageChatCompletionRequestUserMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestUserMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestUserMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestUserMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestUserMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestUserMessageContentArrayOfContentPart], ] @@ -242,14 +410,26 @@ class MessageChatCompletionRequestUserMessage(TypedDict, total=False): information. """ - content: Required[Union[str, SequenceNotStr[MessageChatCompletionRequestUserMessageContentArrayOfContentPart]]] + content: Required[MessageChatCompletionRequestUserMessageContent] """The contents of the user message.""" role: Required[Literal["user"]] """The role of the messages author, in this case `user`.""" -class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnionMember1(TypedDict, total=False): +class MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): """Content part with type and text""" text: Required[str] @@ -258,9 +438,45 @@ class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnion type: Required[Literal["text"]] """The type of content part""" + cache_control: MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart: TypeAlias = Union[ - str, MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartUnionMember1 + str, MessageChatCompletionRequestAssistantMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestAssistantMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestAssistantMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart], ] @@ -296,15 +512,80 @@ class MessageChatCompletionRequestAssistantMessage(TypedDict, total=False): role: Required[Literal["assistant"]] """The role of the messages author, in this case `assistant`.""" - content: Union[str, SequenceNotStr[MessageChatCompletionRequestAssistantMessageContentArrayOfContentPart], None] + content: Optional[MessageChatCompletionRequestAssistantMessageContent] """The contents of the assistant message.""" tool_calls: Iterable[MessageChatCompletionRequestAssistantMessageToolCall] """The tool calls generated by the model, such as function calls.""" +class MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartText(TypedDict, total=False): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartTextCacheControl + """Cache control settings for the content part.""" + + +class MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl( + TypedDict, total=False +): + """Cache control settings for the content part.""" + + type: Required[Literal["ephemeral"]] + """The cache control type.""" + + ttl: Literal["5m", "1h"] + """The cache TTL.""" + + +class MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartText( + TypedDict, total=False +): + """Content part with type and text""" + + text: Required[str] + """The text content""" + + type: Required[Literal["text"]] + """The type of content part""" + + cache_control: ( + MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartTextCacheControl + ) + """Cache control settings for the content part.""" + + +MessageChatCompletionRequestToolMessageContentArrayOfContentPart: TypeAlias = Union[ + str, MessageChatCompletionRequestToolMessageContentArrayOfContentPartChatCompletionRequestContentPartText +] + +MessageChatCompletionRequestToolMessageContent: TypeAlias = Union[ + str, + MessageChatCompletionRequestToolMessageContentChatCompletionRequestContentPartText, + SequenceNotStr[MessageChatCompletionRequestToolMessageContentArrayOfContentPart], +] + + class MessageChatCompletionRequestToolMessage(TypedDict, total=False): - content: Required[str] + content: Required[MessageChatCompletionRequestToolMessageContent] """The contents of the tool message.""" role: Required[Literal["tool"]] diff --git a/src/gradient/types/knowledge_base_create_params.py b/src/gradient/types/knowledge_base_create_params.py index 24cfd98b..2c7bece1 100644 --- a/src/gradient/types/knowledge_base_create_params.py +++ b/src/gradient/types/knowledge_base_create_params.py @@ -3,7 +3,7 @@ from __future__ import annotations from typing import Iterable -from typing_extensions import TypedDict +from typing_extensions import Literal, TypedDict from .._types import SequenceNotStr from .knowledge_bases.aws_data_source_param import AwsDataSourceParam @@ -11,7 +11,13 @@ from .knowledge_bases.api_file_upload_data_source_param import APIFileUploadDataSourceParam from .knowledge_bases.api_web_crawler_data_source_param import APIWebCrawlerDataSourceParam -__all__ = ["KnowledgeBaseCreateParams", "Datasource", "DatasourceDropboxDataSource", "DatasourceGoogleDriveDataSource"] +__all__ = [ + "KnowledgeBaseCreateParams", + "Datasource", + "DatasourceChunkingOptions", + "DatasourceDropboxDataSource", + "DatasourceGoogleDriveDataSource", +] class KnowledgeBaseCreateParams(TypedDict, total=False): @@ -52,6 +58,25 @@ class KnowledgeBaseCreateParams(TypedDict, total=False): """The VPC to deploy the knowledge base database in""" +class DatasourceChunkingOptions(TypedDict, total=False): + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature preview flag.** + """ + + child_chunk_size: int + """Hierarchical options""" + + max_chunk_size: int + """Section_Based and Fixed_Length options""" + + parent_chunk_size: int + """Hierarchical options""" + + semantic_threshold: float + """Semantic options""" + + class DatasourceDropboxDataSource(TypedDict, total=False): """Dropbox Data Source""" @@ -88,6 +113,26 @@ class Datasource(TypedDict, total=False): bucket_region: str """Deprecated, moved to data_source_details""" + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + """The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + + chunking_options: DatasourceChunkingOptions + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + dropbox_data_source: DatasourceDropboxDataSource """Dropbox Data Source""" diff --git a/src/gradient/types/knowledge_bases/__init__.py b/src/gradient/types/knowledge_bases/__init__.py index a8ce2cc7..e7b88be8 100644 --- a/src/gradient/types/knowledge_bases/__init__.py +++ b/src/gradient/types/knowledge_bases/__init__.py @@ -10,12 +10,14 @@ from .indexing_job_list_params import IndexingJobListParams as IndexingJobListParams from .data_source_create_params import DataSourceCreateParams as DataSourceCreateParams from .data_source_list_response import DataSourceListResponse as DataSourceListResponse +from .data_source_update_params import DataSourceUpdateParams as DataSourceUpdateParams from .indexing_job_create_params import IndexingJobCreateParams as IndexingJobCreateParams from .indexing_job_list_response import IndexingJobListResponse as IndexingJobListResponse from .api_file_upload_data_source import APIFileUploadDataSource as APIFileUploadDataSource from .api_web_crawler_data_source import APIWebCrawlerDataSource as APIWebCrawlerDataSource from .data_source_create_response import DataSourceCreateResponse as DataSourceCreateResponse from .data_source_delete_response import DataSourceDeleteResponse as DataSourceDeleteResponse +from .data_source_update_response import DataSourceUpdateResponse as DataSourceUpdateResponse from .api_spaces_data_source_param import APISpacesDataSourceParam as APISpacesDataSourceParam from .indexing_job_create_response import IndexingJobCreateResponse as IndexingJobCreateResponse from .api_knowledge_base_data_source import APIKnowledgeBaseDataSource as APIKnowledgeBaseDataSource diff --git a/src/gradient/types/knowledge_bases/api_knowledge_base_data_source.py b/src/gradient/types/knowledge_bases/api_knowledge_base_data_source.py index b73e325e..ef5db1a5 100644 --- a/src/gradient/types/knowledge_bases/api_knowledge_base_data_source.py +++ b/src/gradient/types/knowledge_bases/api_knowledge_base_data_source.py @@ -2,6 +2,7 @@ from typing import Optional from datetime import datetime +from typing_extensions import Literal from ..._models import BaseModel from .api_spaces_data_source import APISpacesDataSource @@ -9,7 +10,13 @@ from .api_file_upload_data_source import APIFileUploadDataSource from .api_web_crawler_data_source import APIWebCrawlerDataSource -__all__ = ["APIKnowledgeBaseDataSource", "AwsDataSource", "DropboxDataSource", "GoogleDriveDataSource"] +__all__ = [ + "APIKnowledgeBaseDataSource", + "AwsDataSource", + "ChunkingOptions", + "DropboxDataSource", + "GoogleDriveDataSource", +] class AwsDataSource(BaseModel): @@ -24,6 +31,25 @@ class AwsDataSource(BaseModel): """Region of bucket""" +class ChunkingOptions(BaseModel): + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature preview flag.** + """ + + child_chunk_size: Optional[int] = None + """Hierarchical options""" + + max_chunk_size: Optional[int] = None + """Section_Based and Fixed_Length options""" + + parent_chunk_size: Optional[int] = None + """Hierarchical options""" + + semantic_threshold: Optional[float] = None + """Semantic options""" + + class DropboxDataSource(BaseModel): """Dropbox Data Source for Display""" @@ -48,6 +74,28 @@ class APIKnowledgeBaseDataSource(BaseModel): bucket_name: Optional[str] = None """Name of storage bucket - Deprecated, moved to data_source_details""" + chunking_algorithm: Optional[ + Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + ] = None + """The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + + chunking_options: Optional[ChunkingOptions] = None + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + created_at: Optional[datetime] = None """Creation date / time""" diff --git a/src/gradient/types/knowledge_bases/data_source_create_params.py b/src/gradient/types/knowledge_bases/data_source_create_params.py index ac3aa93c..bc65e42a 100644 --- a/src/gradient/types/knowledge_bases/data_source_create_params.py +++ b/src/gradient/types/knowledge_bases/data_source_create_params.py @@ -2,20 +2,40 @@ from __future__ import annotations -from typing_extensions import Annotated, TypedDict +from typing_extensions import Literal, Annotated, TypedDict from ..._utils import PropertyInfo from .aws_data_source_param import AwsDataSourceParam from .api_spaces_data_source_param import APISpacesDataSourceParam from .api_web_crawler_data_source_param import APIWebCrawlerDataSourceParam -__all__ = ["DataSourceCreateParams"] +__all__ = ["DataSourceCreateParams", "ChunkingOptions"] class DataSourceCreateParams(TypedDict, total=False): aws_data_source: AwsDataSourceParam """AWS S3 Data Source""" + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + """The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + + chunking_options: ChunkingOptions + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + body_knowledge_base_uuid: Annotated[str, PropertyInfo(alias="knowledge_base_uuid")] """Knowledge base id""" @@ -24,3 +44,22 @@ class DataSourceCreateParams(TypedDict, total=False): web_crawler_data_source: APIWebCrawlerDataSourceParam """WebCrawlerDataSource""" + + +class ChunkingOptions(TypedDict, total=False): + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature preview flag.** + """ + + child_chunk_size: int + """Hierarchical options""" + + max_chunk_size: int + """Section_Based and Fixed_Length options""" + + parent_chunk_size: int + """Hierarchical options""" + + semantic_threshold: float + """Semantic options""" diff --git a/src/gradient/types/knowledge_bases/data_source_update_params.py b/src/gradient/types/knowledge_bases/data_source_update_params.py new file mode 100644 index 00000000..ffcdf95b --- /dev/null +++ b/src/gradient/types/knowledge_bases/data_source_update_params.py @@ -0,0 +1,58 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from typing_extensions import Literal, Required, Annotated, TypedDict + +from ..._utils import PropertyInfo + +__all__ = ["DataSourceUpdateParams", "ChunkingOptions"] + + +class DataSourceUpdateParams(TypedDict, total=False): + path_knowledge_base_uuid: Required[Annotated[str, PropertyInfo(alias="knowledge_base_uuid")]] + + chunking_algorithm: Literal[ + "CHUNKING_ALGORITHM_UNKNOWN", + "CHUNKING_ALGORITHM_SECTION_BASED", + "CHUNKING_ALGORITHM_HIERARCHICAL", + "CHUNKING_ALGORITHM_SEMANTIC", + "CHUNKING_ALGORITHM_FIXED_LENGTH", + ] + """The chunking algorithm to use for processing data sources. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + + chunking_options: ChunkingOptions + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature + preview flag.** + """ + + body_data_source_uuid: Annotated[str, PropertyInfo(alias="data_source_uuid")] + """Data Source ID (Path Parameter)""" + + body_knowledge_base_uuid: Annotated[str, PropertyInfo(alias="knowledge_base_uuid")] + """Knowledge Base ID (Path Parameter)""" + + +class ChunkingOptions(TypedDict, total=False): + """Configuration options for the chunking algorithm. + + **Note: This feature requires enabling the knowledgebase enhancements feature preview flag.** + """ + + child_chunk_size: int + """Hierarchical options""" + + max_chunk_size: int + """Section_Based and Fixed_Length options""" + + parent_chunk_size: int + """Hierarchical options""" + + semantic_threshold: float + """Semantic options""" diff --git a/src/gradient/types/knowledge_bases/data_source_update_response.py b/src/gradient/types/knowledge_bases/data_source_update_response.py new file mode 100644 index 00000000..31484137 --- /dev/null +++ b/src/gradient/types/knowledge_bases/data_source_update_response.py @@ -0,0 +1,17 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from typing import Optional + +from ..._models import BaseModel +from .api_knowledge_base_data_source import APIKnowledgeBaseDataSource + +__all__ = ["DataSourceUpdateResponse"] + + +class DataSourceUpdateResponse(BaseModel): + """ + Update a data source of a knowledge base with change in chunking algorithm/options + """ + + knowledge_base_data_source: Optional[APIKnowledgeBaseDataSource] = None + """Data Source configuration for Knowledge Bases""" diff --git a/src/gradient/types/response_create_params.py b/src/gradient/types/response_create_params.py new file mode 100644 index 00000000..9b870f3c --- /dev/null +++ b/src/gradient/types/response_create_params.py @@ -0,0 +1,331 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +from typing import Dict, List, Union, Iterable, Optional +from typing_extensions import Literal, Required, TypeAlias, TypedDict + +from .._types import SequenceNotStr + +__all__ = [ + "ResponseCreateParamsBase", + "InputUnionMember1", + "InputUnionMember1UnionMember0", + "InputUnionMember1UnionMember0Content", + "InputUnionMember1UnionMember1", + "InputUnionMember1UnionMember1ContentUnionMember1", + "InputUnionMember1UnionMember1ContentUnionMember1UnionMember0", + "InputUnionMember1UnionMember1ToolCall", + "InputUnionMember1UnionMember1ToolCallFunction", + "StreamOptions", + "ToolChoice", + "ToolChoiceChatCompletionNamedToolChoice", + "ToolChoiceChatCompletionNamedToolChoiceFunction", + "Tool", + "ResponseCreateParamsNonStreaming", + "ResponseCreateParamsStreaming", +] + + +class ResponseCreateParamsBase(TypedDict, total=False): + input: Required[Union[str, Iterable[InputUnionMember1]]] + """The input text prompt or conversation history. + + Can be a string or an array of message objects for conversation context. + """ + + model: Required[str] + """Model ID used to generate the response. Must be a VLLM model.""" + + instructions: Optional[str] + """System-level instructions for the model. + + This sets the behavior and context for the response generation. + """ + + max_output_tokens: Optional[int] + """Maximum number of tokens to generate in the response. + + If not specified, the model will use a default value. + """ + + max_tokens: Optional[int] + """The maximum number of tokens that can be generated in the completion. + + Alias for max_output_tokens for compatibility. + """ + + metadata: Optional[Dict[str, str]] + """Set of 16 key-value pairs that can be attached to an object. + + This can be useful for storing additional information about the object in a + structured format, and querying for objects via API or the dashboard. + + Keys are strings with a maximum length of 64 characters. Values are strings with + a maximum length of 512 characters. + """ + + modalities: Optional[List[Literal["text"]]] + """Specifies the output types the model should generate. + + For text-to-text, this should be ["text"]. + """ + + parallel_tool_calls: Optional[bool] + """Whether to enable parallel tool calls. + + When true, the model can make multiple tool calls in parallel. + """ + + stop: Union[Optional[str], SequenceNotStr[str], None] + """Up to 4 sequences where the API will stop generating further tokens. + + The returned text will not contain the stop sequence. + """ + + stream_options: Optional[StreamOptions] + """Options for streaming response. Only set this when you set `stream: true`.""" + + temperature: Optional[float] + """What sampling temperature to use, between 0 and 2. + + Higher values like 0.8 will make the output more random, while lower values like + 0.2 will make it more focused and deterministic. We generally recommend altering + this or `top_p` but not both. + """ + + tool_choice: ToolChoice + """ + Controls which (if any) tool is called by the model. `none` means the model will + not call any tool and instead generates a message. `auto` means the model can + pick between generating a message or calling one or more tools. `required` means + the model must call one or more tools. Specifying a particular tool via + `{"type": "function", "function": {"name": "my_function"}}` forces the model to + call that tool. + + `none` is the default when no tools are present. `auto` is the default if tools + are present. + """ + + tools: Iterable[Tool] + """A list of tools the model may call. + + Currently, only functions are supported as a tool. Uses Responses API format + (with `name`, `description`, `parameters` at top level). + """ + + top_p: Optional[float] + """ + An alternative to sampling with temperature, called nucleus sampling, where the + model considers the results of the tokens with top_p probability mass. So 0.1 + means only the tokens comprising the top 10% probability mass are considered. + + We generally recommend altering this or `temperature` but not both. + """ + + user: str + """ + A unique identifier representing your end-user, which can help DigitalOcean to + monitor and detect abuse. + """ + + +class InputUnionMember1UnionMember0ContentTyped(TypedDict, total=False): + text: str + """The reasoning text content""" + + type: Literal["reasoning_text"] + """The type of content""" + + +InputUnionMember1UnionMember0Content: TypeAlias = Union[InputUnionMember1UnionMember0ContentTyped, Dict[str, object]] + + +class InputUnionMember1UnionMember0Typed(TypedDict, total=False): + type: Required[Literal["function_call", "function_call_output", "reasoning"]] + """ + The type of input item (must be function_call, function_call_output, or + reasoning) + """ + + id: str + """The unique ID of the reasoning item (optional for reasoning)""" + + arguments: str + """JSON string of function arguments (required for function_call)""" + + call_id: str + """The call ID (required for function_call and function_call_output)""" + + content: Optional[Iterable[InputUnionMember1UnionMember0Content]] + """Array of reasoning content parts (optional for reasoning, can be null)""" + + encrypted_content: Optional[str] + """Encrypted content (optional)""" + + name: str + """The function name (required for function_call)""" + + output: str + """JSON string of function output (required for function_call_output)""" + + status: Optional[str] + """Status of the item (optional, can be null)""" + + summary: Iterable[object] + """Summary of the reasoning (optional for reasoning)""" + + +InputUnionMember1UnionMember0: TypeAlias = Union[InputUnionMember1UnionMember0Typed, Dict[str, object]] + + +class InputUnionMember1UnionMember1ContentUnionMember1UnionMember0(TypedDict, total=False): + text: Required[str] + """The text content""" + + type: Required[Literal["input_text"]] + """The type of content part""" + + +InputUnionMember1UnionMember1ContentUnionMember1: TypeAlias = Union[ + InputUnionMember1UnionMember1ContentUnionMember1UnionMember0, Dict[str, object] +] + + +class InputUnionMember1UnionMember1ToolCallFunction(TypedDict, total=False): + """The function that the model called.""" + + arguments: Required[str] + """ + The arguments to call the function with, as generated by the model in JSON + format. Note that the model does not always generate valid JSON, and may + hallucinate parameters not defined by your function schema. Validate the + arguments in your code before calling your function. + """ + + name: Required[str] + """The name of the function to call.""" + + +class InputUnionMember1UnionMember1ToolCall(TypedDict, total=False): + id: Required[str] + """The ID of the tool call.""" + + function: Required[InputUnionMember1UnionMember1ToolCallFunction] + """The function that the model called.""" + + type: Required[Literal["function"]] + """The type of the tool. Currently, only `function` is supported.""" + + +class InputUnionMember1UnionMember1Typed(TypedDict, total=False): + content: Required[Union[str, Iterable[InputUnionMember1UnionMember1ContentUnionMember1]]] + """The content of the message (string or content parts array)""" + + role: Literal["user", "assistant", "system", "tool", "developer"] + """The role of the message author""" + + tool_call_id: str + """Tool call ID that this message is responding to (required for tool role)""" + + tool_calls: Iterable[InputUnionMember1UnionMember1ToolCall] + """Tool calls made by the assistant (for assistant role messages)""" + + type: Literal["message"] + """Optional type identifier for message items (used by some clients like Codex)""" + + +InputUnionMember1UnionMember1: TypeAlias = Union[InputUnionMember1UnionMember1Typed, Dict[str, object]] + +InputUnionMember1: TypeAlias = Union[InputUnionMember1UnionMember0, InputUnionMember1UnionMember1] + + +class StreamOptions(TypedDict, total=False): + """Options for streaming response. Only set this when you set `stream: true`.""" + + include_usage: bool + """If set, an additional chunk will be streamed before the `data: [DONE]` message. + + The `usage` field on this chunk shows the token usage statistics for the entire + request, and the `choices` field will always be an empty array. + + All other chunks will also include a `usage` field, but with a null value. + **NOTE:** If the stream is interrupted, you may not receive the final usage + chunk which contains the total token usage for the request. + """ + + +class ToolChoiceChatCompletionNamedToolChoiceFunction(TypedDict, total=False): + name: Required[str] + """The name of the function to call.""" + + +class ToolChoiceChatCompletionNamedToolChoice(TypedDict, total=False): + """Specifies a tool the model should use. + + Use to force the model to call a specific function. + """ + + function: Required[ToolChoiceChatCompletionNamedToolChoiceFunction] + + type: Required[Literal["function"]] + """The type of the tool. Currently, only `function` is supported.""" + + +ToolChoice: TypeAlias = Union[Literal["none", "auto", "required"], ToolChoiceChatCompletionNamedToolChoice] + + +class Tool(TypedDict, total=False): + """Tool definition for Responses API (flat format). + + This format is used by VLLM's Responses API where name, description, and parameters are at the top level of the tool object. + """ + + type: Required[Literal["function", "web_search", "web_search_2025_08_26"]] + """The type of the tool. + + Supported values are `function` (custom tools), `web_search`, and + `web_search_2025_08_26` (built-in web search). + """ + + description: str + """ + A description of what the function does, used by the model to choose when and + how to call the function. + """ + + name: str + """The name of the function to be called. + + Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length + of 64. + """ + + parameters: Dict[str, object] + """The parameters the functions accepts, described as a JSON Schema object. + + See the [guide](/docs/guides/function-calling) for examples, and the + [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for + documentation about the format. + + Omitting `parameters` defines a function with an empty parameter list. + """ + + +class ResponseCreateParamsNonStreaming(ResponseCreateParamsBase, total=False): + stream: Optional[Literal[False]] + """ + If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + """ + + +class ResponseCreateParamsStreaming(ResponseCreateParamsBase): + stream: Required[Literal[True]] + """ + If set to true, the model response data will be streamed to the client as it is + generated using server-sent events. + """ + + +ResponseCreateParams = Union[ResponseCreateParamsNonStreaming, ResponseCreateParamsStreaming] diff --git a/src/gradient/types/shared/__init__.py b/src/gradient/types/shared/__init__.py index 4fb2986a..272092b7 100644 --- a/src/gradient/types/shared/__init__.py +++ b/src/gradient/types/shared/__init__.py @@ -26,7 +26,9 @@ from .chat_completion_chunk import ChatCompletionChunk as ChatCompletionChunk from .image_gen_stream_event import ImageGenStreamEvent as ImageGenStreamEvent from .subscription_tier_base import SubscriptionTierBase as SubscriptionTierBase +from .create_response_response import CreateResponseResponse as CreateResponseResponse from .image_gen_completed_event import ImageGenCompletedEvent as ImageGenCompletedEvent from .droplet_next_backup_window import DropletNextBackupWindow as DropletNextBackupWindow from .chat_completion_token_logprob import ChatCompletionTokenLogprob as ChatCompletionTokenLogprob from .image_gen_partial_image_event import ImageGenPartialImageEvent as ImageGenPartialImageEvent +from .create_response_stream_response import CreateResponseStreamResponse as CreateResponseStreamResponse diff --git a/src/gradient/types/shared/completion_usage.py b/src/gradient/types/shared/completion_usage.py index 79ce64ee..596bd045 100644 --- a/src/gradient/types/shared/completion_usage.py +++ b/src/gradient/types/shared/completion_usage.py @@ -2,12 +2,31 @@ from ..._models import BaseModel -__all__ = ["CompletionUsage"] +__all__ = ["CompletionUsage", "CacheCreation"] + + +class CacheCreation(BaseModel): + """Breakdown of prompt tokens written to cache.""" + + ephemeral_1h_input_tokens: int + """Number of prompt tokens written to 1h cache.""" + + ephemeral_5m_input_tokens: int + """Number of prompt tokens written to 5m cache.""" class CompletionUsage(BaseModel): """Usage statistics for the completion request.""" + cache_created_input_tokens: int + """Number of prompt tokens written to cache.""" + + cache_creation: CacheCreation + """Breakdown of prompt tokens written to cache.""" + + cache_read_input_tokens: int + """Number of prompt tokens read from cache.""" + completion_tokens: int """Number of tokens in the generated completion.""" diff --git a/src/gradient/types/shared/create_response_response.py b/src/gradient/types/shared/create_response_response.py new file mode 100644 index 00000000..61dbb155 --- /dev/null +++ b/src/gradient/types/shared/create_response_response.py @@ -0,0 +1,332 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +import builtins +from typing import Dict, List, Union, Optional +from typing_extensions import Literal, TypeAlias + +from ..._models import BaseModel +from .chat_completion_token_logprob import ChatCompletionTokenLogprob + +__all__ = [ + "CreateResponseResponse", + "Usage", + "UsageInputTokensDetails", + "UsageOutputTokensDetails", + "Choice", + "ChoiceMessage", + "ChoiceMessageToolCall", + "ChoiceMessageToolCallFunction", + "ChoiceLogprobs", + "Output", + "OutputUnionMember0", + "OutputUnionMember1", + "OutputUnionMember2", + "OutputUnionMember2Content", + "Tool", +] + + +class UsageInputTokensDetails(BaseModel): + """A detailed breakdown of the input tokens.""" + + cached_tokens: int + """The number of tokens that were retrieved from the cache. + + [More on prompt caching](https://platform.openai.com/docs/guides/prompt-caching). + """ + + +class UsageOutputTokensDetails(BaseModel): + """A detailed breakdown of the output tokens.""" + + reasoning_tokens: int + """The number of reasoning tokens.""" + + tool_output_tokens: int + """The number of tool output tokens.""" + + +class Usage(BaseModel): + """ + Detailed token usage statistics for the request, including input/output token counts and detailed breakdowns. + """ + + input_tokens: int + """The number of input tokens.""" + + input_tokens_details: UsageInputTokensDetails + """A detailed breakdown of the input tokens.""" + + output_tokens: int + """The number of output tokens.""" + + output_tokens_details: UsageOutputTokensDetails + """A detailed breakdown of the output tokens.""" + + total_tokens: int + """The total number of tokens used.""" + + +class ChoiceMessageToolCallFunction(BaseModel): + """The function that the model called.""" + + arguments: str + """ + The arguments to call the function with, as generated by the model in JSON + format. Note that the model does not always generate valid JSON, and may + hallucinate parameters not defined by your function schema. Validate the + arguments in your code before calling your function. + """ + + name: str + """The name of the function to call.""" + + +class ChoiceMessageToolCall(BaseModel): + id: str + """The ID of the tool call.""" + + function: ChoiceMessageToolCallFunction + """The function that the model called.""" + + type: Literal["function"] + """The type of the tool. Currently, only `function` is supported.""" + + +class ChoiceMessage(BaseModel): + """The generated message response.""" + + content: Optional[str] = None + """The generated text content.""" + + role: Optional[Literal["assistant"]] = None + """The role of the message author, which is always `assistant`.""" + + tool_calls: Optional[List[ChoiceMessageToolCall]] = None + """The tool calls generated by the model, such as function calls.""" + + +class ChoiceLogprobs(BaseModel): + """Log probability information for the choice. + + Only present if logprobs was requested in the request. + """ + + content: Optional[List[ChatCompletionTokenLogprob]] = None + """A list of message content tokens with log probability information.""" + + +class Choice(BaseModel): + finish_reason: Literal["stop", "length", "tool_calls", "content_filter"] + """The reason the model stopped generating tokens. + + This will be `stop` if the model hit a natural stop point or a provided stop + sequence, `length` if the maximum number of tokens specified in the request was + reached, or `tool_calls` if the model called a tool. + """ + + index: int + """The index of the choice in the list of choices.""" + + message: ChoiceMessage + """The generated message response.""" + + logprobs: Optional[ChoiceLogprobs] = None + """Log probability information for the choice. + + Only present if logprobs was requested in the request. + """ + + +class OutputUnionMember0(BaseModel): + arguments: str + """JSON string of function arguments""" + + call_id: str + """The unique ID of the function tool call""" + + name: str + """The name of the function to call""" + + type: Literal["function_call"] + """The type of output item""" + + id: Optional[str] = None + """The unique ID of the function tool call (same as call_id)""" + + encrypted_content: Optional[str] = None + """Encrypted content (optional)""" + + status: Optional[str] = None + """Status of the item (optional, can be null)""" + + +class OutputUnionMember1(BaseModel): + text: str + """The text content""" + + type: Literal["text"] + """The type of output item""" + + +class OutputUnionMember2Content(BaseModel): + text: str + """The reasoning text content""" + + type: Literal["reasoning_text"] + """The type of content""" + + +class OutputUnionMember2(BaseModel): + id: str + """The unique ID of the reasoning item""" + + content: List[OutputUnionMember2Content] + """Array of reasoning content parts""" + + summary: List[object] + """Summary of the reasoning (usually empty)""" + + type: Literal["reasoning"] + """The type of output item""" + + encrypted_content: Optional[str] = None + """Encrypted content (optional)""" + + status: Optional[str] = None + """Status of the item (optional, can be null)""" + + +Output: TypeAlias = Union[OutputUnionMember0, OutputUnionMember1, OutputUnionMember2] + + +class Tool(BaseModel): + """Tool definition for Responses API (flat format). + + This format is used by VLLM's Responses API where name, description, and parameters are at the top level of the tool object. + """ + + type: Literal["function", "web_search", "web_search_2025_08_26"] + """The type of the tool. + + Supported values are `function` (custom tools), `web_search`, and + `web_search_2025_08_26` (built-in web search). + """ + + description: Optional[str] = None + """ + A description of what the function does, used by the model to choose when and + how to call the function. + """ + + name: Optional[str] = None + """The name of the function to be called. + + Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length + of 64. + """ + + parameters: Optional[Dict[str, object]] = None + """The parameters the functions accepts, described as a JSON Schema object. + + See the [guide](/docs/guides/function-calling) for examples, and the + [JSON Schema reference](https://json-schema.org/understanding-json-schema/) for + documentation about the format. + + Omitting `parameters` defines a function with an empty parameter list. + """ + + +class CreateResponseResponse(BaseModel): + """ + Represents a text-to-text response returned by the model, based on the provided input. VLLM models only. + """ + + id: str + """A unique identifier for the response.""" + + created: int + """The Unix timestamp (in seconds) of when the response was created.""" + + model: str + """The model used to generate the response.""" + + object: Literal["response"] + """The object type, which is always `response`.""" + + usage: Usage + """ + Detailed token usage statistics for the request, including input/output token + counts and detailed breakdowns. + """ + + background: Optional[bool] = None + """Whether the request was processed in the background""" + + choices: Optional[List[Choice]] = None + """A list of response choices. + + Can be more than one if `n` is greater than 1. Optional - Responses API + primarily uses the output array. + """ + + input_messages: Optional[List[builtins.object]] = None + """Input messages (if applicable)""" + + max_output_tokens: Optional[int] = None + """Maximum output tokens setting""" + + max_tool_calls: Optional[int] = None + """Maximum tool calls setting""" + + output: Optional[List[Output]] = None + """An array of content items generated by the model. + + This includes text content, function calls, reasoning items, and other output + types. Use this field for Responses API compatibility. + """ + + output_messages: Optional[List[builtins.object]] = None + """Output messages (if applicable)""" + + parallel_tool_calls: Optional[bool] = None + """Whether parallel tool calls are enabled""" + + previous_response_id: Optional[str] = None + """Previous response ID (for multi-turn conversations)""" + + prompt: Optional[str] = None + """Prompt used for the response""" + + reasoning: Optional[str] = None + """Reasoning content""" + + service_tier: Optional[str] = None + """Service tier used""" + + status: Optional[str] = None + """Status of the response""" + + temperature: Optional[float] = None + """Temperature setting used for the response""" + + text: Optional[str] = None + """Text content""" + + tool_choice: Optional[str] = None + """Tool choice setting used for the response""" + + tools: Optional[List[Tool]] = None + """Tools available for the response""" + + top_logprobs: Optional[int] = None + """Top logprobs setting""" + + top_p: Optional[float] = None + """Top-p setting used for the response""" + + truncation: Optional[str] = None + """Truncation setting""" + + user: Optional[str] = None + """User identifier""" diff --git a/src/gradient/types/shared/create_response_stream_response.py b/src/gradient/types/shared/create_response_stream_response.py new file mode 100644 index 00000000..ef0230c8 --- /dev/null +++ b/src/gradient/types/shared/create_response_stream_response.py @@ -0,0 +1,139 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from typing import List, Optional +from typing_extensions import Literal + +from ..._models import BaseModel +from .completion_usage import CompletionUsage +from .chat_completion_token_logprob import ChatCompletionTokenLogprob + +__all__ = [ + "CreateResponseStreamResponse", + "Choice", + "ChoiceDelta", + "ChoiceDeltaToolCall", + "ChoiceDeltaToolCallFunction", + "ChoiceLogprobs", +] + + +class ChoiceDeltaToolCallFunction(BaseModel): + """The function that the model called.""" + + arguments: str + """ + The arguments to call the function with, as generated by the model in JSON + format. Note that the model does not always generate valid JSON, and may + hallucinate parameters not defined by your function schema. Validate the + arguments in your code before calling your function. + """ + + name: str + """The name of the function to call.""" + + +class ChoiceDeltaToolCall(BaseModel): + id: str + """The ID of the tool call.""" + + function: ChoiceDeltaToolCallFunction + """The function that the model called.""" + + type: Literal["function"] + """The type of the tool. Currently, only `function` is supported.""" + + +class ChoiceDelta(BaseModel): + """A chunk of the response message generated by the model.""" + + content: Optional[str] = None + """The contents of the chunk message. + + Can be null for chunks with tool calls or other non-text content. + """ + + reasoning_content: Optional[str] = None + """The reasoning content generated by the model. + + Only present when the model generates reasoning text. + """ + + role: Optional[Literal["assistant"]] = None + """The role of the message author. Only present in the first chunk.""" + + tool_calls: Optional[List[ChoiceDeltaToolCall]] = None + """The tool calls generated by the model, such as function calls. + + Only present when the model decides to call a tool. + """ + + +class ChoiceLogprobs(BaseModel): + """Log probability information for the choice. + + Only present if logprobs was requested in the request. + """ + + content: Optional[List[ChatCompletionTokenLogprob]] = None + """A list of message content tokens with log probability information.""" + + +class Choice(BaseModel): + delta: ChoiceDelta + """A chunk of the response message generated by the model.""" + + index: int + """The index of the choice in the list of choices.""" + + finish_reason: Optional[Literal["stop", "length", "tool_calls", "content_filter"]] = None + """The reason the model stopped generating tokens. + + This will be `stop` if the model hit a natural stop point or a provided stop + sequence, `length` if the maximum number of tokens specified in the request was + reached, or `tool_calls` if the model called a tool. Only present in the final + chunk. + """ + + logprobs: Optional[ChoiceLogprobs] = None + """Log probability information for the choice. + + Only present if logprobs was requested in the request. + """ + + +class CreateResponseStreamResponse(BaseModel): + """ + Represents a streamed chunk of a text-to-text response returned by the model, based on the provided input. VLLM models only. + """ + + id: str + """A unique identifier for the response. Each chunk has the same ID.""" + + choices: List[Choice] + """A list of response choice chunks. + + Can contain more than one element if `n` is greater than 1. Can also be empty + for the last chunk if you set `stream_options: {"include_usage": true}`. + """ + + created: int + """The Unix timestamp (in seconds) of when the response was created. + + Each chunk has the same timestamp. + """ + + model: str + """The model used to generate the response.""" + + object: Literal["response.chunk"] + """The object type, which is always `response.chunk`.""" + + usage: Optional[CompletionUsage] = None + """ + An optional field that will only be present when you set + `stream_options: {"include_usage": true}` in your request. When present, it + contains a null value **except for the last chunk** which contains the token + usage statistics for the entire request. **NOTE:** If the stream is interrupted + or cancelled, you may not receive the final usage chunk which contains the total + token usage for the request. + """ diff --git a/tests/api_resources/agents/test_evaluation_datasets.py b/tests/api_resources/agents/test_evaluation_datasets.py index 64dceb03..5093660e 100644 --- a/tests/api_resources/agents/test_evaluation_datasets.py +++ b/tests/api_resources/agents/test_evaluation_datasets.py @@ -30,6 +30,7 @@ def test_method_create(self, client: Gradient) -> None: @parametrize def test_method_create_with_all_params(self, client: Gradient) -> None: evaluation_dataset = client.agents.evaluation_datasets.create( + dataset_type="EVALUATION_DATASET_TYPE_UNKNOWN", file_upload_dataset={ "original_file_name": "example name", "size_in_bytes": "12345", @@ -126,6 +127,7 @@ async def test_method_create(self, async_client: AsyncGradient) -> None: @parametrize async def test_method_create_with_all_params(self, async_client: AsyncGradient) -> None: evaluation_dataset = await async_client.agents.evaluation_datasets.create( + dataset_type="EVALUATION_DATASET_TYPE_UNKNOWN", file_upload_dataset={ "original_file_name": "example name", "size_in_bytes": "12345", diff --git a/tests/api_resources/agents/test_evaluation_runs.py b/tests/api_resources/agents/test_evaluation_runs.py index c6acaf82..8fdfe2cd 100644 --- a/tests/api_resources/agents/test_evaluation_runs.py +++ b/tests/api_resources/agents/test_evaluation_runs.py @@ -32,6 +32,7 @@ def test_method_create(self, client: Gradient) -> None: @parametrize def test_method_create_with_all_params(self, client: Gradient) -> None: evaluation_run = client.agents.evaluation_runs.create( + agent_deployment_names=["example string"], agent_uuids=["example string"], run_name="Evaluation Run Name", test_case_uuid='"12345678-1234-1234-1234-123456789012"', @@ -216,6 +217,7 @@ async def test_method_create(self, async_client: AsyncGradient) -> None: @parametrize async def test_method_create_with_all_params(self, async_client: AsyncGradient) -> None: evaluation_run = await async_client.agents.evaluation_runs.create( + agent_deployment_names=["example string"], agent_uuids=["example string"], run_name="Evaluation Run Name", test_case_uuid='"12345678-1234-1234-1234-123456789012"', diff --git a/tests/api_resources/agents/test_evaluation_test_cases.py b/tests/api_resources/agents/test_evaluation_test_cases.py index 7cd0a07e..a8942239 100644 --- a/tests/api_resources/agents/test_evaluation_test_cases.py +++ b/tests/api_resources/agents/test_evaluation_test_cases.py @@ -33,6 +33,7 @@ def test_method_create(self, client: Gradient) -> None: @parametrize def test_method_create_with_all_params(self, client: Gradient) -> None: evaluation_test_case = client.agents.evaluation_test_cases.create( + agent_workspace_name="example name", dataset_uuid="123e4567-e89b-12d3-a456-426614174000", description="example string", metrics=["example string"], @@ -278,6 +279,7 @@ async def test_method_create(self, async_client: AsyncGradient) -> None: @parametrize async def test_method_create_with_all_params(self, async_client: AsyncGradient) -> None: evaluation_test_case = await async_client.agents.evaluation_test_cases.create( + agent_workspace_name="example name", dataset_uuid="123e4567-e89b-12d3-a456-426614174000", description="example string", metrics=["example string"], diff --git a/tests/api_resources/apps/__init__.py b/tests/api_resources/apps/__init__.py new file mode 100644 index 00000000..fd8019a9 --- /dev/null +++ b/tests/api_resources/apps/__init__.py @@ -0,0 +1 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. diff --git a/tests/api_resources/apps/test_job_invocations.py b/tests/api_resources/apps/test_job_invocations.py new file mode 100644 index 00000000..a7cb68df --- /dev/null +++ b/tests/api_resources/apps/test_job_invocations.py @@ -0,0 +1,148 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +import os +from typing import Any, cast + +import pytest + +from gradient import Gradient, AsyncGradient +from tests.utils import assert_matches_type +from gradient.types.apps import JobInvocationCancelResponse + +base_url = os.environ.get("TEST_API_BASE_URL", "http://127.0.0.1:4010") + + +class TestJobInvocations: + parametrize = pytest.mark.parametrize("client", [False, True], indirect=True, ids=["loose", "strict"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_cancel(self, client: Gradient) -> None: + job_invocation = client.apps.job_invocations.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_cancel_with_all_params(self, client: Gradient) -> None: + job_invocation = client.apps.job_invocations.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + job_name="job_name", + ) + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_raw_response_cancel(self, client: Gradient) -> None: + response = client.apps.job_invocations.with_raw_response.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) + + assert response.is_closed is True + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + job_invocation = response.parse() + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_streaming_response_cancel(self, client: Gradient) -> None: + with client.apps.job_invocations.with_streaming_response.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + job_invocation = response.parse() + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + assert cast(Any, response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_path_params_cancel(self, client: Gradient) -> None: + with pytest.raises(ValueError, match=r"Expected a non-empty value for `app_id` but received ''"): + client.apps.job_invocations.with_raw_response.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="", + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `job_invocation_id` but received ''"): + client.apps.job_invocations.with_raw_response.cancel( + job_invocation_id="", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) + + +class TestAsyncJobInvocations: + parametrize = pytest.mark.parametrize( + "async_client", [False, True, {"http_client": "aiohttp"}], indirect=True, ids=["loose", "strict", "aiohttp"] + ) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_cancel(self, async_client: AsyncGradient) -> None: + job_invocation = await async_client.apps.job_invocations.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_cancel_with_all_params(self, async_client: AsyncGradient) -> None: + job_invocation = await async_client.apps.job_invocations.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + job_name="job_name", + ) + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_raw_response_cancel(self, async_client: AsyncGradient) -> None: + response = await async_client.apps.job_invocations.with_raw_response.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) + + assert response.is_closed is True + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + job_invocation = await response.parse() + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_streaming_response_cancel(self, async_client: AsyncGradient) -> None: + async with async_client.apps.job_invocations.with_streaming_response.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + job_invocation = await response.parse() + assert_matches_type(JobInvocationCancelResponse, job_invocation, path=["response"]) + + assert cast(Any, response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_path_params_cancel(self, async_client: AsyncGradient) -> None: + with pytest.raises(ValueError, match=r"Expected a non-empty value for `app_id` but received ''"): + await async_client.apps.job_invocations.with_raw_response.cancel( + job_invocation_id="123e4567-e89b-12d3-a456-426", + app_id="", + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `job_invocation_id` but received ''"): + await async_client.apps.job_invocations.with_raw_response.cancel( + job_invocation_id="", + app_id="4f6c71e2-1e90-4762-9fee-6cc4a0a9f2cf", + ) diff --git a/tests/api_resources/gpu_droplets/test_images.py b/tests/api_resources/gpu_droplets/test_images.py index 4c4146e2..480f94a5 100644 --- a/tests/api_resources/gpu_droplets/test_images.py +++ b/tests/api_resources/gpu_droplets/test_images.py @@ -32,9 +32,9 @@ def test_method_create(self, client: Gradient) -> None: @parametrize def test_method_create_with_all_params(self, client: Gradient) -> None: image = client.gpu_droplets.images.create( - description=" ", + description="Cloud-optimized image w/ small footprint", distribution="Ubuntu", - name="Nifty New Snapshot", + name="ubuntu-18.04-minimal", region="nyc3", tags=["base-image", "prod"], url="http://cloud-images.ubuntu.com/minimal/releases/bionic/release/ubuntu-18.04-minimal-cloudimg-amd64.img", @@ -232,9 +232,9 @@ async def test_method_create(self, async_client: AsyncGradient) -> None: @parametrize async def test_method_create_with_all_params(self, async_client: AsyncGradient) -> None: image = await async_client.gpu_droplets.images.create( - description=" ", + description="Cloud-optimized image w/ small footprint", distribution="Ubuntu", - name="Nifty New Snapshot", + name="ubuntu-18.04-minimal", region="nyc3", tags=["base-image", "prod"], url="http://cloud-images.ubuntu.com/minimal/releases/bionic/release/ubuntu-18.04-minimal-cloudimg-amd64.img", diff --git a/tests/api_resources/knowledge_bases/test_data_sources.py b/tests/api_resources/knowledge_bases/test_data_sources.py index ca721d93..d28fd409 100644 --- a/tests/api_resources/knowledge_bases/test_data_sources.py +++ b/tests/api_resources/knowledge_bases/test_data_sources.py @@ -13,6 +13,7 @@ DataSourceListResponse, DataSourceCreateResponse, DataSourceDeleteResponse, + DataSourceUpdateResponse, DataSourceCreatePresignedURLsResponse, ) @@ -42,6 +43,13 @@ def test_method_create_with_all_params(self, client: Gradient) -> None: "region": "example string", "secret_key": "example string", }, + chunking_algorithm="CHUNKING_ALGORITHM_SECTION_BASED", + chunking_options={ + "child_chunk_size": 350, + "max_chunk_size": 750, + "parent_chunk_size": 1000, + "semantic_threshold": 0.5, + }, body_knowledge_base_uuid='"12345678-1234-1234-1234-123456789012"', spaces_data_source={ "bucket_name": "example name", @@ -93,6 +101,78 @@ def test_path_params_create(self, client: Gradient) -> None: path_knowledge_base_uuid="", ) + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_update(self, client: Gradient) -> None: + data_source = client.knowledge_bases.data_sources.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_update_with_all_params(self, client: Gradient) -> None: + data_source = client.knowledge_bases.data_sources.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + chunking_algorithm="CHUNKING_ALGORITHM_SECTION_BASED", + chunking_options={ + "child_chunk_size": 350, + "max_chunk_size": 750, + "parent_chunk_size": 1000, + "semantic_threshold": 0.5, + }, + body_data_source_uuid="98765432-1234-1234-1234-123456789012", + body_knowledge_base_uuid="12345678-1234-1234-1234-123456789012", + ) + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_raw_response_update(self, client: Gradient) -> None: + response = client.knowledge_bases.data_sources.with_raw_response.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) + + assert response.is_closed is True + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + data_source = response.parse() + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_streaming_response_update(self, client: Gradient) -> None: + with client.knowledge_bases.data_sources.with_streaming_response.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + data_source = response.parse() + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + assert cast(Any, response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_path_params_update(self, client: Gradient) -> None: + with pytest.raises( + ValueError, match=r"Expected a non-empty value for `path_knowledge_base_uuid` but received ''" + ): + client.knowledge_bases.data_sources.with_raw_response.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="", + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `path_data_source_uuid` but received ''"): + client.knowledge_bases.data_sources.with_raw_response.update( + path_data_source_uuid="", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) + @pytest.mark.skip(reason="Prism tests are disabled") @parametrize def test_method_list(self, client: Gradient) -> None: @@ -264,6 +344,13 @@ async def test_method_create_with_all_params(self, async_client: AsyncGradient) "region": "example string", "secret_key": "example string", }, + chunking_algorithm="CHUNKING_ALGORITHM_SECTION_BASED", + chunking_options={ + "child_chunk_size": 350, + "max_chunk_size": 750, + "parent_chunk_size": 1000, + "semantic_threshold": 0.5, + }, body_knowledge_base_uuid='"12345678-1234-1234-1234-123456789012"', spaces_data_source={ "bucket_name": "example name", @@ -315,6 +402,78 @@ async def test_path_params_create(self, async_client: AsyncGradient) -> None: path_knowledge_base_uuid="", ) + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_update(self, async_client: AsyncGradient) -> None: + data_source = await async_client.knowledge_bases.data_sources.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_update_with_all_params(self, async_client: AsyncGradient) -> None: + data_source = await async_client.knowledge_bases.data_sources.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + chunking_algorithm="CHUNKING_ALGORITHM_SECTION_BASED", + chunking_options={ + "child_chunk_size": 350, + "max_chunk_size": 750, + "parent_chunk_size": 1000, + "semantic_threshold": 0.5, + }, + body_data_source_uuid="98765432-1234-1234-1234-123456789012", + body_knowledge_base_uuid="12345678-1234-1234-1234-123456789012", + ) + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_raw_response_update(self, async_client: AsyncGradient) -> None: + response = await async_client.knowledge_bases.data_sources.with_raw_response.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) + + assert response.is_closed is True + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + data_source = await response.parse() + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_streaming_response_update(self, async_client: AsyncGradient) -> None: + async with async_client.knowledge_bases.data_sources.with_streaming_response.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + data_source = await response.parse() + assert_matches_type(DataSourceUpdateResponse, data_source, path=["response"]) + + assert cast(Any, response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_path_params_update(self, async_client: AsyncGradient) -> None: + with pytest.raises( + ValueError, match=r"Expected a non-empty value for `path_knowledge_base_uuid` but received ''" + ): + await async_client.knowledge_bases.data_sources.with_raw_response.update( + path_data_source_uuid="123e4567-e89b-12d3-a456-426614174000", + path_knowledge_base_uuid="", + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `path_data_source_uuid` but received ''"): + await async_client.knowledge_bases.data_sources.with_raw_response.update( + path_data_source_uuid="", + path_knowledge_base_uuid="123e4567-e89b-12d3-a456-426614174000", + ) + @pytest.mark.skip(reason="Prism tests are disabled") @parametrize async def test_method_list(self, async_client: AsyncGradient) -> None: diff --git a/tests/api_resources/test_billing.py b/tests/api_resources/test_billing.py new file mode 100644 index 00000000..59181b15 --- /dev/null +++ b/tests/api_resources/test_billing.py @@ -0,0 +1,177 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +import os +from typing import Any, cast + +import pytest + +from gradient import Gradient, AsyncGradient +from tests.utils import assert_matches_type +from gradient.types import BillingListInsightsResponse +from gradient._utils import parse_date + +base_url = os.environ.get("TEST_API_BASE_URL", "http://127.0.0.1:4010") + + +class TestBilling: + parametrize = pytest.mark.parametrize("client", [False, True], indirect=True, ids=["loose", "strict"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_list_insights(self, client: Gradient) -> None: + billing = client.billing.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_list_insights_with_all_params(self, client: Gradient) -> None: + billing = client.billing.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + page=1, + per_page=1, + ) + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_raw_response_list_insights(self, client: Gradient) -> None: + response = client.billing.with_raw_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) + + assert response.is_closed is True + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + billing = response.parse() + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_streaming_response_list_insights(self, client: Gradient) -> None: + with client.billing.with_streaming_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + billing = response.parse() + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + assert cast(Any, response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_path_params_list_insights(self, client: Gradient) -> None: + with pytest.raises(ValueError, match=r"Expected a non-empty value for `account_urn` but received ''"): + client.billing.with_raw_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="", + start_date=parse_date("2025-01-01"), + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `start_date` but received ''"): + client.billing.with_raw_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date="", + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `end_date` but received ''"): + client.billing.with_raw_response.list_insights( + end_date="", + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) + + +class TestAsyncBilling: + parametrize = pytest.mark.parametrize( + "async_client", [False, True, {"http_client": "aiohttp"}], indirect=True, ids=["loose", "strict", "aiohttp"] + ) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_list_insights(self, async_client: AsyncGradient) -> None: + billing = await async_client.billing.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_list_insights_with_all_params(self, async_client: AsyncGradient) -> None: + billing = await async_client.billing.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + page=1, + per_page=1, + ) + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_raw_response_list_insights(self, async_client: AsyncGradient) -> None: + response = await async_client.billing.with_raw_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) + + assert response.is_closed is True + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + billing = await response.parse() + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_streaming_response_list_insights(self, async_client: AsyncGradient) -> None: + async with async_client.billing.with_streaming_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + billing = await response.parse() + assert_matches_type(BillingListInsightsResponse, billing, path=["response"]) + + assert cast(Any, response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_path_params_list_insights(self, async_client: AsyncGradient) -> None: + with pytest.raises(ValueError, match=r"Expected a non-empty value for `account_urn` but received ''"): + await async_client.billing.with_raw_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="", + start_date=parse_date("2025-01-01"), + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `start_date` but received ''"): + await async_client.billing.with_raw_response.list_insights( + end_date=parse_date("2025-01-31"), + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date="", + ) + + with pytest.raises(ValueError, match=r"Expected a non-empty value for `end_date` but received ''"): + await async_client.billing.with_raw_response.list_insights( + end_date="", + account_urn="do:team:12345678-1234-1234-1234-123456789012", + start_date=parse_date("2025-01-01"), + ) diff --git a/tests/api_resources/test_knowledge_bases.py b/tests/api_resources/test_knowledge_bases.py index 632951b4..9ce9785d 100644 --- a/tests/api_resources/test_knowledge_bases.py +++ b/tests/api_resources/test_knowledge_bases.py @@ -46,6 +46,13 @@ def test_method_create_with_all_params(self, client: Gradient) -> None: }, "bucket_name": "example name", "bucket_region": "example string", + "chunking_algorithm": "CHUNKING_ALGORITHM_SECTION_BASED", + "chunking_options": { + "child_chunk_size": 350, + "max_chunk_size": 750, + "parent_chunk_size": 1000, + "semantic_threshold": 0.5, + }, "dropbox_data_source": { "folder": "example string", "refresh_token": "example string", @@ -447,6 +454,13 @@ async def test_method_create_with_all_params(self, async_client: AsyncGradient) }, "bucket_name": "example name", "bucket_region": "example string", + "chunking_algorithm": "CHUNKING_ALGORITHM_SECTION_BASED", + "chunking_options": { + "child_chunk_size": 350, + "max_chunk_size": 750, + "parent_chunk_size": 1000, + "semantic_threshold": 0.5, + }, "dropbox_data_source": { "folder": "example string", "refresh_token": "example string", diff --git a/tests/api_resources/test_responses.py b/tests/api_resources/test_responses.py new file mode 100644 index 00000000..0d8d7acf --- /dev/null +++ b/tests/api_resources/test_responses.py @@ -0,0 +1,296 @@ +# File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. + +from __future__ import annotations + +import os +from typing import Any, cast + +import pytest + +from gradient import Gradient, AsyncGradient +from tests.utils import assert_matches_type +from gradient.types.shared import CreateResponseResponse + +base_url = os.environ.get("TEST_API_BASE_URL", "http://127.0.0.1:4010") + + +class TestResponses: + parametrize = pytest.mark.parametrize("client", [False, True], indirect=True, ids=["loose", "strict"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_create_overload_1(self, client: Gradient) -> None: + response = client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + ) + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_create_with_all_params_overload_1(self, client: Gradient) -> None: + response = client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + instructions="You are a helpful assistant.", + max_output_tokens=1024, + max_tokens=1024, + metadata={"foo": "string"}, + modalities=["text"], + parallel_tool_calls=True, + stop="\n", + stream=False, + stream_options={"include_usage": True}, + temperature=1, + tool_choice="none", + tools=[ + { + "type": "function", + "description": "description", + "name": "name", + "parameters": {"foo": "bar"}, + } + ], + top_p=1, + user="user-1234", + ) + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_raw_response_create_overload_1(self, client: Gradient) -> None: + http_response = client.responses.with_raw_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + ) + + assert http_response.is_closed is True + assert http_response.http_request.headers.get("X-Stainless-Lang") == "python" + response = http_response.parse() + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_streaming_response_create_overload_1(self, client: Gradient) -> None: + with client.responses.with_streaming_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + ) as http_response: + assert not http_response.is_closed + assert http_response.http_request.headers.get("X-Stainless-Lang") == "python" + + response = http_response.parse() + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + assert cast(Any, http_response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_create_overload_2(self, client: Gradient) -> None: + response_stream = client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + ) + response_stream.response.close() + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_method_create_with_all_params_overload_2(self, client: Gradient) -> None: + response_stream = client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + instructions="You are a helpful assistant.", + max_output_tokens=1024, + max_tokens=1024, + metadata={"foo": "string"}, + modalities=["text"], + parallel_tool_calls=True, + stop="\n", + stream_options={"include_usage": True}, + temperature=1, + tool_choice="none", + tools=[ + { + "type": "function", + "description": "description", + "name": "name", + "parameters": {"foo": "bar"}, + } + ], + top_p=1, + user="user-1234", + ) + response_stream.response.close() + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_raw_response_create_overload_2(self, client: Gradient) -> None: + response = client.responses.with_raw_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + ) + + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + stream = response.parse() + stream.close() + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + def test_streaming_response_create_overload_2(self, client: Gradient) -> None: + with client.responses.with_streaming_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + stream = response.parse() + stream.close() + + assert cast(Any, response.is_closed) is True + + +class TestAsyncResponses: + parametrize = pytest.mark.parametrize( + "async_client", [False, True, {"http_client": "aiohttp"}], indirect=True, ids=["loose", "strict", "aiohttp"] + ) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_create_overload_1(self, async_client: AsyncGradient) -> None: + response = await async_client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + ) + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_create_with_all_params_overload_1(self, async_client: AsyncGradient) -> None: + response = await async_client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + instructions="You are a helpful assistant.", + max_output_tokens=1024, + max_tokens=1024, + metadata={"foo": "string"}, + modalities=["text"], + parallel_tool_calls=True, + stop="\n", + stream=False, + stream_options={"include_usage": True}, + temperature=1, + tool_choice="none", + tools=[ + { + "type": "function", + "description": "description", + "name": "name", + "parameters": {"foo": "bar"}, + } + ], + top_p=1, + user="user-1234", + ) + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_raw_response_create_overload_1(self, async_client: AsyncGradient) -> None: + http_response = await async_client.responses.with_raw_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + ) + + assert http_response.is_closed is True + assert http_response.http_request.headers.get("X-Stainless-Lang") == "python" + response = await http_response.parse() + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_streaming_response_create_overload_1(self, async_client: AsyncGradient) -> None: + async with async_client.responses.with_streaming_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + ) as http_response: + assert not http_response.is_closed + assert http_response.http_request.headers.get("X-Stainless-Lang") == "python" + + response = await http_response.parse() + assert_matches_type(CreateResponseResponse, response, path=["response"]) + + assert cast(Any, http_response.is_closed) is True + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_create_overload_2(self, async_client: AsyncGradient) -> None: + response_stream = await async_client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + ) + await response_stream.response.aclose() + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_method_create_with_all_params_overload_2(self, async_client: AsyncGradient) -> None: + response_stream = await async_client.responses.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + instructions="You are a helpful assistant.", + max_output_tokens=1024, + max_tokens=1024, + metadata={"foo": "string"}, + modalities=["text"], + parallel_tool_calls=True, + stop="\n", + stream_options={"include_usage": True}, + temperature=1, + tool_choice="none", + tools=[ + { + "type": "function", + "description": "description", + "name": "name", + "parameters": {"foo": "bar"}, + } + ], + top_p=1, + user="user-1234", + ) + await response_stream.response.aclose() + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_raw_response_create_overload_2(self, async_client: AsyncGradient) -> None: + response = await async_client.responses.with_raw_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + ) + + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + stream = await response.parse() + await stream.close() + + @pytest.mark.skip(reason="Prism tests are disabled") + @parametrize + async def test_streaming_response_create_overload_2(self, async_client: AsyncGradient) -> None: + async with async_client.responses.with_streaming_response.create( + input="Tell me a three-sentence bedtime story about a unicorn.", + model="llama3-8b-instruct", + stream=True, + ) as response: + assert not response.is_closed + assert response.http_request.headers.get("X-Stainless-Lang") == "python" + + stream = await response.parse() + await stream.close() + + assert cast(Any, response.is_closed) is True diff --git a/tests/test_client.py b/tests/test_client.py index c78fa9f8..d5f1bbe6 100644 --- a/tests/test_client.py +++ b/tests/test_client.py @@ -8,10 +8,11 @@ import json import asyncio import inspect +import dataclasses import tracemalloc -from typing import Any, Union, cast +from typing import Any, Union, TypeVar, Callable, Iterable, Iterator, Optional, Coroutine, cast from unittest import mock -from typing_extensions import Literal +from typing_extensions import Literal, AsyncIterator, override import httpx import pytest @@ -41,6 +42,7 @@ from .utils import update_env +T = TypeVar("T") base_url = os.environ.get("TEST_API_BASE_URL", "http://127.0.0.1:4010") access_token = "My Access Token" model_access_key = "My Model Access Key" @@ -57,6 +59,57 @@ def _low_retry_timeout(*_args: Any, **_kwargs: Any) -> float: return 0.1 +def mirror_request_content(request: httpx.Request) -> httpx.Response: + return httpx.Response(200, content=request.content) + + +# note: we can't use the httpx.MockTransport class as it consumes the request +# body itself, which means we can't test that the body is read lazily +class MockTransport(httpx.BaseTransport, httpx.AsyncBaseTransport): + def __init__( + self, + handler: Callable[[httpx.Request], httpx.Response] + | Callable[[httpx.Request], Coroutine[Any, Any, httpx.Response]], + ) -> None: + self.handler = handler + + @override + def handle_request( + self, + request: httpx.Request, + ) -> httpx.Response: + assert not inspect.iscoroutinefunction(self.handler), "handler must not be a coroutine function" + assert inspect.isfunction(self.handler), "handler must be a function" + return self.handler(request) + + @override + async def handle_async_request( + self, + request: httpx.Request, + ) -> httpx.Response: + assert inspect.iscoroutinefunction(self.handler), "handler must be a coroutine function" + return await self.handler(request) + + +@dataclasses.dataclass +class Counter: + value: int = 0 + + +def _make_sync_iterator(iterable: Iterable[T], counter: Optional[Counter] = None) -> Iterator[T]: + for item in iterable: + if counter: + counter.value += 1 + yield item + + +async def _make_async_iterator(iterable: Iterable[T], counter: Optional[Counter] = None) -> AsyncIterator[T]: + for item in iterable: + if counter: + counter.value += 1 + yield item + + def _get_open_connections(client: Gradient | AsyncGradient) -> int: transport = client._client._transport assert isinstance(transport, httpx.HTTPTransport) or isinstance(transport, httpx.AsyncHTTPTransport) @@ -607,6 +660,70 @@ def test_multipart_repeating_array(self, client: Gradient) -> None: b"", ] + @pytest.mark.respx(base_url=base_url) + def test_binary_content_upload(self, respx_mock: MockRouter, client: Gradient) -> None: + respx_mock.post("/upload").mock(side_effect=mirror_request_content) + + file_content = b"Hello, this is a test file." + + response = client.post( + "/upload", + content=file_content, + cast_to=httpx.Response, + options={"headers": {"Content-Type": "application/octet-stream"}}, + ) + + assert response.status_code == 200 + assert response.request.headers["Content-Type"] == "application/octet-stream" + assert response.content == file_content + + def test_binary_content_upload_with_iterator(self) -> None: + file_content = b"Hello, this is a test file." + counter = Counter() + iterator = _make_sync_iterator([file_content], counter=counter) + + def mock_handler(request: httpx.Request) -> httpx.Response: + assert counter.value == 0, "the request body should not have been read" + return httpx.Response(200, content=request.read()) + + with Gradient( + base_url=base_url, + access_token=access_token, + _strict_response_validation=True, + http_client=httpx.Client(transport=MockTransport(handler=mock_handler)), + ) as client: + response = client.post( + "/upload", + content=iterator, + cast_to=httpx.Response, + options={"headers": {"Content-Type": "application/octet-stream"}}, + ) + + assert response.status_code == 200 + assert response.request.headers["Content-Type"] == "application/octet-stream" + assert response.content == file_content + assert counter.value == 1 + + @pytest.mark.respx(base_url=base_url) + def test_binary_content_upload_with_body_is_deprecated(self, respx_mock: MockRouter, client: Gradient) -> None: + respx_mock.post("/upload").mock(side_effect=mirror_request_content) + + file_content = b"Hello, this is a test file." + + with pytest.deprecated_call( + match="Passing raw bytes as `body` is deprecated and will be removed in a future version. Please pass raw bytes via the `content` parameter instead." + ): + response = client.post( + "/upload", + body=file_content, + cast_to=httpx.Response, + options={"headers": {"Content-Type": "application/octet-stream"}}, + ) + + assert response.status_code == 200 + assert response.request.headers["Content-Type"] == "application/octet-stream" + assert response.content == file_content + @pytest.mark.respx(base_url=base_url) def test_basic_union_response(self, respx_mock: MockRouter, client: Gradient) -> None: class Model1(BaseModel): @@ -1648,6 +1765,72 @@ def test_multipart_repeating_array(self, async_client: AsyncGradient) -> None: b"", ] + @pytest.mark.respx(base_url=base_url) + async def test_binary_content_upload(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: + respx_mock.post("/upload").mock(side_effect=mirror_request_content) + + file_content = b"Hello, this is a test file." + + response = await async_client.post( + "/upload", + content=file_content, + cast_to=httpx.Response, + options={"headers": {"Content-Type": "application/octet-stream"}}, + ) + + assert response.status_code == 200 + assert response.request.headers["Content-Type"] == "application/octet-stream" + assert response.content == file_content + + async def test_binary_content_upload_with_asynciterator(self) -> None: + file_content = b"Hello, this is a test file." + counter = Counter() + iterator = _make_async_iterator([file_content], counter=counter) + + async def mock_handler(request: httpx.Request) -> httpx.Response: + assert counter.value == 0, "the request body should not have been read" + return httpx.Response(200, content=await request.aread()) + + async with AsyncGradient( + base_url=base_url, + access_token=access_token, + _strict_response_validation=True, + http_client=httpx.AsyncClient(transport=MockTransport(handler=mock_handler)), + ) as client: + response = await client.post( + "/upload", + content=iterator, + cast_to=httpx.Response, + options={"headers": {"Content-Type": "application/octet-stream"}}, + ) + + assert response.status_code == 200 + assert response.request.headers["Content-Type"] == "application/octet-stream" + assert response.content == file_content + assert counter.value == 1 + + @pytest.mark.respx(base_url=base_url) + async def test_binary_content_upload_with_body_is_deprecated( + self, respx_mock: MockRouter, async_client: AsyncGradient + ) -> None: + respx_mock.post("/upload").mock(side_effect=mirror_request_content) + + file_content = b"Hello, this is a test file." + + with pytest.deprecated_call( + match="Passing raw bytes as `body` is deprecated and will be removed in a future version. Please pass raw bytes via the `content` parameter instead." + ): + response = await async_client.post( + "/upload", + body=file_content, + cast_to=httpx.Response, + options={"headers": {"Content-Type": "application/octet-stream"}}, + ) + + assert response.status_code == 200 + assert response.request.headers["Content-Type"] == "application/octet-stream" + assert response.content == file_content + @pytest.mark.respx(base_url=base_url) async def test_basic_union_response(self, respx_mock: MockRouter, async_client: AsyncGradient) -> None: class Model1(BaseModel): diff --git a/tests/test_utils/test_json.py b/tests/test_utils/test_json.py new file mode 100644 index 00000000..4ba6d83c --- /dev/null +++ b/tests/test_utils/test_json.py @@ -0,0 +1,126 @@ +from __future__ import annotations + +import datetime +from typing import Union + +import pydantic + +from gradient import _compat +from gradient._utils._json import openapi_dumps + + +class TestOpenapiDumps: + def test_basic(self) -> None: + data = {"key": "value", "number": 42} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"key":"value","number":42}' + + def test_datetime_serialization(self) -> None: + dt = datetime.datetime(2023, 1, 1, 12, 0, 0) + data = {"datetime": dt} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"datetime":"2023-01-01T12:00:00"}' + + def test_pydantic_model_serialization(self) -> None: + class User(pydantic.BaseModel): + first_name: str + last_name: str + age: int + + model_instance = User(first_name="John", last_name="Kramer", age=83) + data = {"model": model_instance} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"first_name":"John","last_name":"Kramer","age":83}}' + + def test_pydantic_model_with_default_values(self) -> None: + class User(pydantic.BaseModel): + name: str + role: str = "user" + active: bool = True + score: int = 0 + + model_instance = User(name="Alice") + data = {"model": model_instance} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"name":"Alice"}}' + + def test_pydantic_model_with_default_values_overridden(self) -> None: + class User(pydantic.BaseModel): + name: str + role: str = "user" + active: bool = True + + model_instance = User(name="Bob", role="admin", active=False) + data = {"model": model_instance} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"name":"Bob","role":"admin","active":false}}' + + def test_pydantic_model_with_alias(self) -> None: + class User(pydantic.BaseModel): + first_name: str = pydantic.Field(alias="firstName") + last_name: str = pydantic.Field(alias="lastName") + + model_instance = User(firstName="John", lastName="Doe") + data = {"model": model_instance} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"firstName":"John","lastName":"Doe"}}' + + def test_pydantic_model_with_alias_and_default(self) -> None: + class User(pydantic.BaseModel): + user_name: str = pydantic.Field(alias="userName") + user_role: str = pydantic.Field(default="member", alias="userRole") + is_active: bool = pydantic.Field(default=True, alias="isActive") + + model_instance = User(userName="charlie") + data = {"model": model_instance} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"userName":"charlie"}}' + + model_with_overrides = User(userName="diana", userRole="admin", isActive=False) + data = {"model": model_with_overrides} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"userName":"diana","userRole":"admin","isActive":false}}' + + def test_pydantic_model_with_nested_models_and_defaults(self) -> None: + class Address(pydantic.BaseModel): + street: str + city: str = "Unknown" + + class User(pydantic.BaseModel): + name: str + address: Address + verified: bool = False + + if _compat.PYDANTIC_V1: + # to handle forward references in Pydantic v1 + User.update_forward_refs(**locals()) # type: ignore[reportDeprecated] + + address = Address(street="123 Main St") + user = User(name="Diana", address=address) + data = {"user": user} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"user":{"name":"Diana","address":{"street":"123 Main St"}}}' + + address_with_city = Address(street="456 Oak Ave", city="Boston") + user_verified = User(name="Eve", address=address_with_city, verified=True) + data = {"user": user_verified} + json_bytes = openapi_dumps(data) + assert ( + json_bytes == b'{"user":{"name":"Eve","address":{"street":"456 Oak Ave","city":"Boston"},"verified":true}}' + ) + + def test_pydantic_model_with_optional_fields(self) -> None: + class User(pydantic.BaseModel): + name: str + email: Union[str, None] + phone: Union[str, None] + + model_with_none = User(name="Eve", email=None, phone=None) + data = {"model": model_with_none} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"name":"Eve","email":null,"phone":null}}' + + model_with_values = User(name="Frank", email="frank@example.com", phone=None) + data = {"model": model_with_values} + json_bytes = openapi_dumps(data) + assert json_bytes == b'{"model":{"name":"Frank","email":"frank@example.com","phone":null}}' From 26ac7960527278e89cce6694f2d880f7e26e9c94 Mon Sep 17 00:00:00 2001 From: rushitat Date: Wed, 11 Feb 2026 15:46:31 -0800 Subject: [PATCH 4/8] Fix lint --- src/gradient/_base_client.py | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/src/gradient/_base_client.py b/src/gradient/_base_client.py index e2135952..ca3db359 100644 --- a/src/gradient/_base_client.py +++ b/src/gradient/_base_client.py @@ -62,7 +62,7 @@ not_given, ) from ._utils import is_dict, is_list, asyncify, is_given, lru_cache, is_mapping -from ._compat import PYDANTIC_V1, model_copy, model_dump +from ._compat import PYDANTIC_V1, model_copy from ._models import GenericModel, FinalRequestOptions, validate_type, construct_type from ._response import ( APIResponse, @@ -486,17 +486,17 @@ def _build_request( ) -> httpx.Request: if log.isEnabledFor(logging.DEBUG): log.debug( - "Request options: %s", - model_dump( - options, - exclude_unset=True, - # Pydantic v1 can't dump every type we support in content, so we exclude it for now. - exclude={ - "content", - } - if PYDANTIC_V1 - else {}, - ), + "Request options", + # model_dump( + # options, + # exclude_unset=True, + # # Pydantic v1 can't dump every type we support in content, so we exclude it for now. + # exclude={ + # "content", + # } + # if PYDANTIC_V1 + # else {}, + # ), ) kwargs: dict[str, Any] = {} From 21520fdac633e7bcc1be4ddbbd6f214e5a624484 Mon Sep 17 00:00:00 2001 From: rushitat Date: Wed, 11 Feb 2026 22:09:46 -0800 Subject: [PATCH 5/8] responses --- src/gradient/resources/responses.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/gradient/resources/responses.py b/src/gradient/resources/responses.py index a342a78d..d0892fa9 100644 --- a/src/gradient/resources/responses.py +++ b/src/gradient/resources/responses.py @@ -392,7 +392,7 @@ def create( timeout: float | httpx.Timeout | None | NotGiven = not_given, ) -> CreateResponseResponse | Stream[CreateResponseStreamResponse]: return self._post( - "/responses", + "/responses" if self._client._base_url_overridden else f"{self._client.inference_endpoint}/v1/responses", body=maybe_transform( { "input": input, @@ -791,7 +791,7 @@ async def create( timeout: float | httpx.Timeout | None | NotGiven = not_given, ) -> CreateResponseResponse | AsyncStream[CreateResponseStreamResponse]: return await self._post( - "/responses", + "/responses" if self._client._base_url_overridden else f"{self._client.inference_endpoint}/v1/responses", body=await async_maybe_transform( { "input": input, From 9c18bdbbec7e7f8222099e9ada5780b61cde22fa Mon Sep 17 00:00:00 2001 From: rushitat Date: Thu, 12 Feb 2026 08:47:37 -0800 Subject: [PATCH 6/8] Add TYPE_CHECKING imports for GPUDropletsResource in _client --- src/gradient/_client.py | 1 + 1 file changed, 1 insertion(+) diff --git a/src/gradient/_client.py b/src/gradient/_client.py index 18ef0c7a..a0fc87b2 100644 --- a/src/gradient/_client.py +++ b/src/gradient/_client.py @@ -64,6 +64,7 @@ KnowledgeBasesResource, AsyncKnowledgeBasesResource, ) + from .resources.gpu_droplets import GPUDropletsResource, AsyncGPUDropletsResource __all__ = [ "Timeout", From 7160d350f24837b987d807ea58dad7c0b63e2ec1 Mon Sep 17 00:00:00 2001 From: rushitat Date: Thu, 12 Feb 2026 08:50:35 -0800 Subject: [PATCH 7/8] Fix Lint --- src/gradient/_client.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/gradient/_client.py b/src/gradient/_client.py index a0fc87b2..42867c7d 100644 --- a/src/gradient/_client.py +++ b/src/gradient/_client.py @@ -56,6 +56,7 @@ from .resources.apps.apps import AppsResource, AsyncAppsResource from .resources.chat.chat import ChatResource, AsyncChatResource from .resources.responses import ResponsesResource, AsyncResponsesResource + from .resources.gpu_droplets import GPUDropletsResource, AsyncGPUDropletsResource from .resources.agents.agents import AgentsResource, AsyncAgentsResource from .resources.models.models import ModelsResource, AsyncModelsResource from .resources.databases.databases import DatabasesResource, AsyncDatabasesResource @@ -64,7 +65,6 @@ KnowledgeBasesResource, AsyncKnowledgeBasesResource, ) - from .resources.gpu_droplets import GPUDropletsResource, AsyncGPUDropletsResource __all__ = [ "Timeout", From 2674257bfacabef90178f53726c74185bba0acbe Mon Sep 17 00:00:00 2001 From: "stainless-app[bot]" <142633134+stainless-app[bot]@users.noreply.github.com> Date: Tue, 17 Feb 2026 18:29:26 +0000 Subject: [PATCH 8/8] release: 3.11.0 --- .release-please-manifest.json | 2 +- CHANGELOG.md | 18 ++++++++++++++++++ pyproject.toml | 2 +- src/gradient/_version.py | 2 +- 4 files changed, 21 insertions(+), 3 deletions(-) diff --git a/.release-please-manifest.json b/.release-please-manifest.json index cc4da81f..940f2ca3 100644 --- a/.release-please-manifest.json +++ b/.release-please-manifest.json @@ -1,3 +1,3 @@ { - ".": "3.10.1" + ".": "3.11.0" } \ No newline at end of file diff --git a/CHANGELOG.md b/CHANGELOG.md index a82ede75..e9caf7e3 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,5 +1,23 @@ # Changelog +## 3.11.0 (2026-02-17) + +Full Changelog: [v3.10.1...v3.11.0](https://github.com/digitalocean/gradient-python/compare/v3.10.1...v3.11.0) + +### Features + +* **api:** api update ([f67228f](https://github.com/digitalocean/gradient-python/commit/f67228f0bd8de96342d15b1b7c3096e4b03c7aaa)) + + +### Bug Fixes + +* **client:** loosen auth header validation ([bf5fa9f](https://github.com/digitalocean/gradient-python/commit/bf5fa9f93b6443ac0b41d5fa134aab8ee12cfbc3)) + + +### Chores + +* **internal:** codegen related update ([07eeda8](https://github.com/digitalocean/gradient-python/commit/07eeda8ae3e1c7bdfe3fcc51a8786f35682f4c3f)) + ## 3.10.1 (2025-12-19) Full Changelog: [v3.10.0...v3.10.1](https://github.com/digitalocean/gradient-python/compare/v3.10.0...v3.10.1) diff --git a/pyproject.toml b/pyproject.toml index 5d4e9ebe..3295b4c4 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -1,6 +1,6 @@ [project] name = "gradient" -version = "3.10.1" +version = "3.11.0" description = "The official Python library for the Gradient API" dynamic = ["readme"] license = "Apache-2.0" diff --git a/src/gradient/_version.py b/src/gradient/_version.py index f1eb0021..07c427f5 100644 --- a/src/gradient/_version.py +++ b/src/gradient/_version.py @@ -1,4 +1,4 @@ # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details. __title__ = "gradient" -__version__ = "3.10.1" # x-release-please-version +__version__ = "3.11.0" # x-release-please-version